*Article* **Research-by-Design in Complex Systems: Reflections on Approaches Used to Reimagine Environmentally Sustainable, High-Welfare Poultry Housing Futures**

**Emma Campbell 1,\*, Greg Keeffe 1, Seán Cullen 1, Anne Richmond 2, Stephen Beagan 2, Ursula Lavery 2, Brendan McKenna <sup>3</sup> and Steven Lester <sup>3</sup>**


**Abstract:** Despite projected global rises in chicken consumption, growing environmental and welfare challenges threaten the future of commercial poultry production. Though some of these challenges, such as biosecurity, sourcing, pollution, and waste, have been thoroughly researched, the open-ended, complex, and interrelated nature of the sector means that it is difficult for poultry producers to know how to change. Design may offer a new way to analyse and reframe these challenges, to speculate on a range of different solutions for these complex systems of production. This paper reflects on the research-by-design methods applied to reimagine environmentally sustainable, highwelfare poultry housing futures. The paper is based on an eighteen-month long, multidisciplinary research project with a large U.K.-based poultry farming integrator, a poultry house ventilation and equipment supplier, and academic partners with expertise in research-by-design and bird welfare. After contextualising challenges faced by the poultry sector, the paper outlines a three-step, iterative approach within which design methods were applied, beginning with (1) a baseline analysis of farm inputs, outputs, actors, and networks, and then (2) a consolidation of themes and scenarios, leading to the development of (3) a compendium of ideas for the future of poultry farming. The Results section presents three design propositions, each imagining different futures by recreating the farm as a system of "closed-loop" flows, reframing the "chicken as client" and challenging current centralised models of production to connect consumers to food provenance and impact. These propositions function as vehicles to test design methods, such as designing for resource flows challenging actor hierarchies and hacking stakeholder networks. While some interesting ideas are presented, the paper highlights the complexity of the challenge and reflects on the value of design to reframe these challenges to collaboratively foster new perspectives and mindsets.

**Keywords:** research-by-design methods; design for environmental sustainability; high-welfare farming design; poultry housing futures; visualisation methods

#### **1. Introduction**

By 2030, global meat production is projected to increase by 13%, while poultry production is set to rise by 17% in the same period [1]. Although the sector is relatively low carbon compared with other livestock sectors, poultry production significantly impacts water and air quality and contributes to global resource use and waste [2,3]. In the United Kingdom (U.K.), climate legislation, including net-zero targets, requires the poultry sector to fundamentally redress operations to reduce environmental impact and build resilience to the direct and indirect impacts of climate change. Consumers are also becoming more aware of bird welfare but are reluctant to pay more for their food [1]. As such, higher welfare products still represent a small portion of the chicken market. Poultry producers

**Citation:** Campbell, E.; Keeffe, G.; Cullen, S.; Richmond, A.; Beagan, S.; Lavery, U.; McKenna, B.; Lester, S. Research-by-Design in Complex Systems: Reflections on Approaches Used to Reimagine Environmentally Sustainable, High-Welfare Poultry Housing Futures. *Sustainability* **2023**, *15*, 5808. https://doi.org/10.3390/ su15075808

Academic Editors: Santosh Jagtap and Lucia Corsini

Received: 16 December 2022 Revised: 10 March 2023 Accepted: 21 March 2023 Published: 27 March 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

must balance these concerns along with biosecurity risks that are easier to manage through indoor housing environments. While these challenges have been explored extensively in a siloed way, their collective complexity means that it is difficult to consider how the sector might holistically adapt for the future [3]. This paper reflects on an 18-month long, applied, multidisciplinary project between the poultry sector in Northern Ireland (N.I.) and architectural design-researchers in academia to explore this question through research-by-design methods. As well as the design-research team, the project partners included Moy Park, a large U.K. (United Kingdom) poultry integrator, and JF McKenna, a poultry house ventilation and equipment supplier, and academic experts in poultry behaviour and welfare. The project was funded by Innovate UK to address and support innovation on net-zero and high-welfare agriculture through the design of future poultry housing. However, this paper is reframed towards environmental sustainability, including biosecurity, resource consumption, waste, and pollution issues.

After highlighting the challenges faced by the poultry sector and exploring typical poultry housing models, this paper reflects on the value of research-by-design methods to reimagine environmentally sustainable, high-welfare poultry house futures. It also reflects on how these methods facilitate new perspectives and mindsets within complex productive industries, in the agrifood sector and beyond. The benefit of undertaking this research in an academic setting, across an extended time-period, was in the distancing of focus from the day-to-day operations of the poultry integrator towards the exploration of longand short-term opportunities that enable "more good" rather than attempting to simply make the poultry housing and systems "less bad". The design-research team developed an "Ideation Hourglass" framework to support top-down and bottom-up ideas, spanning different scales and timeframes to inform pathways to change. Under this framework, the team used a three-step approach to understand and reimagine poultry housing. Within each iterative, interconnected step, a range of established design tools were tested, and visualisations produced to capture, communicate, and critique ideas. The cross-disciplinary team met regularly to assess developed ideas and ideate new ones. The project culminated in the development of three propositional concepts for future poultry housing. While these propositions consolidated baseline analysis, themes, and potential scenarios developed to consider poultry sector futures, in this paper, they are used as vehicles to explore and reflect on the mix of design approaches to develop them and how these might inform ways to use design to unpack and reimagine other complex systems.

#### **2. Background**

#### *2.1. Poultry Production Impacts and Challenges*

Around 60 billion chickens are slaughtered globally per year [4]. Agriculture accounts for 14.5% of global anthropogenic greenhouse gas (GHG) emissions [5]. Poultry and egg production contributes to 8% of this figure, around 0.12% of total GHG emissions [5]. In June 2020, 1.15 billion chickens were slaughtered in the U.K., with a total production value of £2.25 billion [6]. In this context, agriculture accounts for 10% of annual greenhouse gas emissions (GHG) [7]. The poultry sector accounts for about 13% of the U.K.'s gross agricultural output and 2% of the agriculture sector's global warming potential [3]. Greenhouse gas emissions in the poultry sector are attributed to a range of operational sources. The global warming potential of standard poultry production systems in the U.K. are associated with feed and water (71.2%), electricity (3.7%), gas and oil (9.8%), housing and land (12.1%) and manure and bedding (3.2%) [3].

In its Sixth Carbon Budget, the Climate Change Committee recommended that the agricultural sector increase efficiencies in resource production, promote biodiversity, and facilitate greenhouse gas removals to address climate targets [8]. In 2022, Northern Ireland (N.I.) agreed on its first Climate Change Bill [9]. Like the UK, it seeks to meet net-zero carbon by 2050 [10]. Despite this, the challenge of making the poultry sector more environmentally sustainable is particularly acute in N.I. as agriculture is the only sector to have increased emissions since the 1990 baseline and livestock farming contributes significantly to the regional economy [11]. The poultry sector makes up 18.5% of N.I.'s gross agricultural output, 5.5% more than the U.K. [11]. Annual greenhouse gas emissions from agriculture accounts for 26% of total emissions, 16% more than the UK [12]. Part of this can be attributed to the rapid increase in the size of the poultry sector, particularly in poultry populations for meat consumption. Between 2016 and 2020, the poultry population in N.I. rose from 14.5 to 15.4 million [13].

The rapid development of this sector has created a complex supply chain and fostered the development of supporting industries, from poultry feed manufacturers to anaerobic digestors for poultry litter management. Furthermore, land spreading is currently one of the primary ways of managing litter from broiler houses, contributing significantly to a phosphorous (P) surplus in N.I. In 2017, the national P surplus was 12.3 kg P ha−<sup>1</sup> [14]. Consequently, eutrophication of waterways is a prominent issue, detrimentally affecting the water quality of major rivers and lakes. In 2021, none of N.I.'s 450 river water bodies achieved good or high-level status, down from 24% in 2015, compared with 14% of English rivers rated good status [15,16]. These pressures have brought into sharp focus the environmental impact of the poultry sector and have opened questions around how best to decarbonise operations and reduce the impacts of sourcing and waste management.

Welfare is another concern for the poultry sector. Most consumers say that it is an important consideration when they purchase meat [17]. Despite this, products labelled as higher welfare, such as organic options, make up a small proportion of market share, reflecting that consumers do not want to spend more on high-welfare products. Commercial poultry producers must also balance bird welfare with biosecurity requirements, to mitigate against seasonal outbreaks of avian influenza (AI). A record number of cases of AI were confirmed in Great Britain (GB) in the winter of 2021/22 [18]. AI affects wild and commercial bird populations and presents a significant public health threat, and once detected on farms, farmers must cull tens of thousands of birds, impacting income as a result [19]. In large-scale poultry production, indoor housing is preferred, despite the impact on welfare [20]. The Better Chicken Commitment outlines ways to improve welfare in indoor-reared birds; by reducing stocking densities, careful breed selection and improved environmental standards, such as natural light, perching spaces, and good air quality [21].

#### *2.2. Poultry House Typology*

Historically, poultry houses were simple, lean-to timber structures capable of holding 10–12 birds. These free-range systems enabled the birds to roam extensively and required the farmer to move long distances to clean or supply feed and water to the houses. More intensive systems emerged, designed to hold hundreds, then thousands of birds at a time. These structures were timber framed, used natural ventilation, and integrated automated feed and drink lines. In recent years, conventional poultry houses have been designed as steel, portal frame structures with insulated façade panels and an uninsulated concrete floor.

Today, Moy Park's average house is approximately 20 m by 80 m and capable of holding 34,500 birds. Each house completes 6.8 cycles per year on average, so that around 234,600 birds occupy a house annually. Heating systems in Moy Park's estate use primarily biomass, with a small proportion using natural gas. As well as being artificially lit, manual opening strip windows with external shutters are installed along the length of the house to provide controlled natural light. The house is ventilated through side inlets and ridge extracts with artificial fan systems to aid air movement. This lightweight structure, with low thermal mass, is vulnerable to the increased occurrence of extreme hot and cold weather in the U.K. [22]. As a result, poultry houses rely heavily on mechanical systems to heat and move air. This is costly, wasteful, and can sometimes be insufficient at maintaining an optimum indoor environment during extreme weather events.

In recent years, notable examples of broiler and laying houses have emerged which aim to reduce operational energy use and improve bird welfare. These deviate from the conventional portal frame typology. The Rondeel is a circular laying house with alternate day and night segments and an outer edge for play and dustbathing [23]. Even though the cost of operating the Rondeel is greater than a typical house, the design significantly improves animal welfare because it hosts smaller flock sizes and creates variable environments to support a range of needs. The Windstreek Broiler House is another example of a new poultry housing typology [24,25]. Its 11-m-high roof avails of natural cross ventilation, while heated motherhoods reduce space heating to several occupied zones. This house has a large north-facing elevation, which results in a significant increase in natural light, with glazing equivalent to 50% of the floor area compared to 3% in conventional housing. These design strategies help to reduce operational energy of the house by 80% when compared to a conventional house. While both designs offer interesting insights into the future of the poultry house, their operation is predicated on designing architectural and management solutions without explicitly integrating novel, circular systems that balance environmental and welfare concerns.

#### **3. Literature and Methods**

Design-research uses abductive reasoning to define complex problems and generate solutions simultaneously and across iterations in what philosopher Karl Popper defines as conjecture and refutation [26]. This is reflected in IDEO's divergent–convergent creative design process [27]. While traditional research is mostly analytical, exploring "what is", research-by-design casts into the future to explore "what if" [28]. This approach, combined with systems thinking, which supports pattern-finding in complex environments, is useful in determining solutions for "wicked" problems, which are difficult to define, incomplete and interdependent [29].

In this research study, the design-research team implicitly referred to Schön's model of Reflective Practice, which defines three stages in a design cycle: frame, move and evaluate [30]. This is mirrored in the structure of this section of the paper, which defines three steps to reimagine and visualise environmentally sustainable, welfare-friendly poultry house futures using research-by-design methods. While there are three steps described, these approaches are iterative and non-linear, with interlinks across and between.

The first step in this research-by-design approach begins with a baseline analysis of Moy Park's existing poultry housing, networks, and operations. While this step is not necessarily design-led, it highlights challenges in which the team can develop design solutions. Keeffe and Cullen describe this step as defining content in which a formal space can be contained [31]. The second step comprises a compendium of themes and scenarios to explore the wider context in which the current and future poultry house exists. The second step straddles analysis and design, to inform "what-if" questions emerging from the research in steps one and two. The third step is design-led, exploring various holistic, thematic solutions across different scales and timeframes.

To balance the mix of expertise in the project, the design-research team developed an "Ideation Hourglass" framework, shown in Figure 1, to facilitate top-down and bottom-up systems-thinking. Where Moy Park and JF McKenna brought in-depth knowledge of current systems and a focus on the incremental changes towards the project's aims, the academic team brought fresh perspectives contextualised by wider societal shifts towards radical change. The project team met regularly to discuss ideas spanning different scales and timeframes. Discussions were playful and informal, aimed at instilling cross-disciplinary ideation between the industry and academic partners to support new perspectives and mindsets about the current and future state of the poultry sector.

Within this framework, several established design methods were tested, such as the REAP method, which supports bottom-up solutions to reduce, reuse, and produce resources across different scales, and the STEEP method, which aids top-down ideation through social, technological ecological/environmental, economic, and political lens-based analysis [32–34]. During the design phase, the COCD-box method was also used to categorise ideas into how, wow, and now, aiding the team to focus on the development of just a few ideas across the spectrum of feasible, not yet feasible, common, and original [35]. Different types of

visualisations, such as diagrams and montages, were used as tools to communicate design ideas and illustrate design methods.

**Figure 1.** "Ideation Hourglass" framework diagram, balancing bottom-up and top-down thinking to design across timescales.

Here, the challenge of reimagining the poultry house of the future was also used to test research-by-design methods to address complex systems of production by designing spaces based on what flows through them, challenging priorities for building users by shaking up actor hierarchies and redesigning spaces based on new stakeholder dynamics and networks.

Each step in this section informs the development of design propositions, presented in the results, which also function as vehicles to reflect on the research-by-design approaches employed. Here the three design outputs are described through the design approaches used to create them.

#### *3.1. Baseline of Typical Farm Inputs, Outputs, Actors and Networks*

The design-research team started the project by carrying out baseline analysis of typical farm spaces and operations, including resource consumption and production. The team also analysed the actors and networks of current operational infrastructures in Moy Park.

#### 3.1.1. Assessing Embodied and Operational Carbon

Assessments of embodied and operation carbon were carried out to understand the environmental impact of poultry housing in the context of the net-zero challenge, a stipulation of the grant funding. First, the team analysed the embodied carbon associated with the construction of a typical poultry house. Through analysis of architectural drawings, and conversations with Moy Park and JF McKenna, the design-research team estimated the quantity of each construction material used in a typical house. Using the ICE embodied

carbon database [36], the quantity of embodied carbon per kg of each material type in the base (aggregate, concrete), primary structure (steel frame, timber purlins), walls (façade panels, concrete, polystyrene insulation, double-glazed windows), and roof (tin sheeting, insulation, PVC membrane) was estimated. The exercise revealed that a typical house embodies around 57.3 tonnes of CO2-eq (tCO2) and found that the steel structure of a typical house was particularly carbon-intensive material, using just under half of this carbon, at 25 tCO2-eq per house.

The team also looked at the operational carbon required to heat and power a typical poultry house each year. Due to time-constraints, the team gathered a breadth of rough data on a typical house. For example, analysis revealed that 90% of all houses across the estate used biomass heating, and 75% of those used wood pellets; therefore, the team carried out carbon assessments based on wood pellet biomass heating. Research revealed that approximately 434,000 kWh is required to heat a house per year. The team found that houses heated by wood pellets produce 15 g CO2-eq per kWh and around 6.5 tonnes of CO2-eq per year [37]. Researchers found that all houses are powered by N.I.'s grid electricity, which uses around 339 g CO2-eq per kWh, and each house uses approximately 34,000 kWh of electricity per year, equivalent to 10.5 homes in N.I., producing 11.5 tonnes of CO2-eq per year [38,39].

Determining that half of embodied carbon in the house corresponded to its primary structure opened reflections on how changing this material to, for example, a timber structure that has lower embodied carbon, could be an easy way to dramatically reduce the embodied carbon in house construction to address net-zero targets. Since the embodied carbon of the primary structure is integral to the house, Moy Park would have to wait until an existing house reaches its end-of-life or adopt lower carbon materials in new houses first. In contrast, the team agreed that it would be easier and quicker to address operational carbon by, for example, adapting the fabric of existing houses to increase thermal mass, to insulate floors and roofs to use and waste less heat. The team also reflected on precedent typologies, mentioned in Section 2.2, to consider more efficient heating and power infrastructures and whether they could be applied in N.I. Dismantling space to focus on carbon, energy or material flows supported these reflections and unlocked new ways of designing the house to rewire these flows. The authors pick up on this point in Section 4.1 of the Section 4.

#### 3.1.2. Assessing Stakeholder Actor-Networks

The design-research team unpacked and visualised other flows, from infrastructural to social, through actor-network mapping. In addition to the carbon assessment of current operations, explicit and tentative relationships between Moy Park and their stakeholders were analysed. Data were collected though informal interviews with sector specialists inside Moy Park's organisation, such as employees working on the management of litter, plastic packaging, and public relations. This informed further desktop analysis of important suppliers, consumers, influencers, and governors.

Modelled on Thun et. al's interpretation of Latour's theory, the actor-network map in Figure 2 was produced to visualise key actors and relationships between them [40]. These were used to inform conversations with Moy Park about how these networks might be manipulated to address the aims of the research project. This method of drawing helped to codify a complex and dynamic web of connections between internal and external stakeholders, providing a system lens for strategic-level conversations around present and future operations.

While Moy Park's internal business structures were clear to employees, relationships with external stakeholder networks were not. As a poultry integrator, the relationship between business and farmer is complex and intertwined, and as such, essential to understand to reimagine future relationships and spaces. One interesting aspect of the poultry integrator's business infrastructure was that the poultry farmers supplying the integrator operated a bit like franchisees, personally purchasing all the necessary ingredients from

the integrator to make fully-grown chickens, such as paying for housing, power, feed, and chicks. To determine the boundaries of environmental responsibility, the team asked the organization questions such as, Who owns the chickens? Who buys the feed? Who is responsible for litter management? Who monitors emissions? Visualisations of the whole network and key relationships in that network helped Moy Park to reflect on previously unseen connections for the first time. This opened conversations on partnerships, sharing, and responsibility on issues of resource consumption, waste, and welfare to consider how the house design might respond to or influence these concerns and highlight opportunities to hack these hierarchies to address the project's aims. This approach was taken in the development of the second design proposition presented in Section 4.2 of the Results.

**Figure 2.** Actor-network map capturing explicit and tentative relationships within Moy Park's business ecosystem.

#### *3.2. Consolidation of Themes and Scenarios*

Subthemes in relation to environmental sustainability and bird welfare emerged as a result of the baseline to encourage more holistic inquiry and prompt further research to understand impacts and opportunities for Moy Park. This encouraged the industry partners to see existing conditions through new eyes allowing them to question these in a safe, explorative environment.

The design-research team developed long-term risk and opportunity scenarios using the STEEP scenario-building approach to explore what is happening in societies, including businesses, at local and global scale [32,33]. This informed the development of five themes of interest to the poultry sector. These themes, described briefly below, formed the basis of informal discussions and desktop research across the project:


Theme-based analysis through scenario-building enabled the design-research team to develop a deeper understanding of core risks in achieving the aims of the project. These were communicated to Moy Park through short descriptions and icon-style diagrams to relay how production operations might be affected in the long-term. These diagrams informed a mnemonic approach, consolidating a collection of complex concerns and ideas into easily understood chunks of information. Compiling key challenges in one, easy-toread format gave the poultry producer a discussion tool for strategic level decision-making beyond the project period. A selection of potential scenarios and mitigations is outlined in Table 1.


**Table 1.** Exploring thematic scenarios and mitigations, supported by icon diagrams.

Parallels between themes revealed the complex nature of transitioning the sector to address environmental sustainability and welfare issues, as well as the impact of prioritising some themes over others, and how this might result in radically different future conditions. The research team compared complementary and conflicting themes to combine problems and solutions. These problem-solution framings are described next, ordered in relation to each of the three design propositions presented in the Results section:


Compromises of one theme over another under the broader context of addressing environmental sustainability and welfare revealed different challenges for different actors within the STEEP nexus. For example, new housing infrastructures require economic investment from individual farmers, which influences the speed of change, while social and political influences could also play a role in catalysing or stalling these changes from the outside. While the STEEP analysis highlighted conflicting challenges, bearing in mind the project's aims, the research team prioritised the ecological/environmental aspect of the STEEP tool as they developed design propositions. In this way, they could employ design approaches to explore alternative systems, spaces, and networks of production unbound from the priorities of current operations. This approach of hacking stakeholder networks is explored further in Section 4.3 of the Results.

#### *3.3. Imagining a Compendium of Multi-Scalar Ideas for the Poultry Farm of the Future*

The design-research team used the themes, scenarios, and baseline findings to build a compendium of propositional ideas and solutions to imagine environmentally sustainable, higher welfare poultry farms and housing. The compendium of future scenarios, spaces, and ideas developed organically, often evolving from unstructured conversations. In these instances, problem-framing through "what-if" questions led to thought-experiments: *"If this happened, then what might be the result? How might this impact society, the environment, or economics? If we removed economic considerations, what decisions might be made, and could the environment or society benefit more?"*. From here, the team applied Cross' concept of a creative leap, finding "sub-solutions" to bridge against the framed problem through conceptual thinking [42]. Significant ideas stemming from conversations were recorded and then were informally framed as a research question. This question was then explored by one team member through more rigorous design-based investigations and presented to the design team on a weekly basis for feedback to inform new iterations. This approach

worked well, giving time to collaboratively bounce ideas through conversation as well as ensuring progress by formalising one or two of these ideas alone. Working in this way, each individual design-researcher brought different value systems, interests, knowledge profiles, and experiences to their propositional outputs. Presenting these formalised ideas in a group setting challenged these value systems, providing positive criticism through different lenses. This meant that one idea, explored by several individuals, would result in different outputs, and therefore more holistic solutions.

Through the "Ideation Hourglass" framework, the project team balanced top-down and bottom-up actions. Where the top-down approach focused on risks and opportunities as well as external pressures, the bottom-up approach focused on immediate action to identify "low-hanging fruit", or operations that could be easily transitioned to environmentally sustainable or higher welfare alternatives. Aided by the baseline analysis findings, the design-research team used the REAP design method to categorise potential "easy wins" at the house, farm, and estate scale in Figure 3 [34]. Taking a mindset that it is better to fix a little bit of the whole system than perfect just one aspect of it, one change applied at each scale could have a multiplier effect to quickly reduce carbon emissions, waste, or reliance on external supply chains.

**Figure 3.** REAP method matrix for categorising bottom-up design ideation.

As research-by-design is iterative, some propositional ideas emerged earlier in the project than others, for example, reviewing typical annual energy use to power a poultry house prompted testing of the potential energy production if solar panels were to be installed onto its roof. Equally, some ideas were developed further than others, while others remained abstract or focused on a certain aspect of the project's aims. In addition, as is typical in design-research, evaluative thinking ensured that problems and solutions co-evolved across the project. For example, development of the propositional ideas incited further baseline analysis as well as further theme definition and scenario-building [26]. Themes placed at the centre of the divergent–convergent thinking process funneled the development of further ideas, while also helping the design-research team to focus on developing just a few ideas more comprehensively. The COCD box method was also used to facilitate critique and design decisions to develop the three propositions [35].

In all instances, each relevant idea was recorded and communicated through a range of drawing formats including diagrams, maps, timelines, collages, orthographic drawings, and sketches with one example provided in Figure 4. The design-research team applied a familiar visualisation approach that combines the production of diagrams and photomontage-style collages to communicate core design ideas and give a sense of how these ideas play out spatially [43]. Read together, both drawing types communicate complex ideas holistically. Read in context, they capture views of an imagined scenario combining existing and future worlds. Drawing across the formats described prompted on-demand decision making: "*How big is the future house in its landscape? How does light enter the space? What type of food is the chicken consuming?*". Once complete, the image supports further reflection on the type of space imagined and whether it effectively addresses the challenges. Like the scenarios described in the previous section, this approach worked as another visual mnemonic to explore many solutions to various parts of a complex design problem. Further examples of the different drawings produced are shown in the Results section of this paper.

**Figure 4.** Sketches depicting early ideas on embedding circularity within the poultry house design and wider farm operations.

#### **4. Results**

The research-by-design methods applied across this project aimed to generate a way of looking at current and reimagining future poultry house architectures to address environmental sustainability and bird welfare in poultry production. Here, the challenges faced by the poultry sector are tested, with the overarching aim of finding design approaches to examine and unlock pathways to reimagine complex locked-in systems of production.

As indicated in the previous section of this paper, the last step in the approach informed the development of a compendium of design propositions to address these aims. Following the generation of five key themes in Section 3.2, a compendium of disparate thematic design ideas was synthesised to the design and visualisation of just three propositional futures for poultry production, explored in this section. Each proposition emerged from lens-based investigation of two of the five themes identified in Section 3.2, as indicated in Figure 5**,** to inform house designs with different priorities relating to the overarching challenges.

The propositions, named *Circular Chicken, Happy Chicken,* and *Network Chicken*, are described below—less as finished projects or optimum visions for the future of poultry housing, and more as vehicles to describe the different research-by-design methods employed in the project to address its overall aims. These propositions are described under the umbrella of the design approaches employed in the project, for example, the *Circu-* *lar Chicken* proposition highlights how the designers shifted their focus from space to flow design.

**Figure 5.** Icon images indicating how different themes, described in Section 3.2, were combined to generate different scenarios for the future of poultry farming and housing.

#### *4.1. Dismantling Space to Flows*

Baseline analysis of the typical poultry house, highlighted in Section 3.1.1, revealed the carbon emissions associated with construction materials, heating, and power supply. This revealed that the constant stream of resources used to operate the house outweighed the one-time impact of constructing the architectural artefact, and as a result, dismantled the initial view that, to achieve net-zero, the house must first be redesigned. This highlighted the need to redesign the house's flows rather than its spaces.

The difficultly of reducing emissions associated with operating the house since, for example, Moy Park is unable to decarbonise external heat and power supply, was also highlighted. This opened conversations around how they might use existing waste streams within their estate to localise heat and power supply, for example, through anerobic digestion of poultry litter. The team also discussed how Moy Park might take advantage of the large roofscape on existing poultry houses to install solar PV panels to build some resilience against the rising price of heat and power supply. The team reviewed the types of technologies required to support this shift, such as battery storage, and how Moy Park might share excess energy production through cooperatives with residents in rural N.I. Further baseline research on the impact of waste outputs, such as poultry litter, revealed unseen environmental damage, which informed conversations around how to promote more sustainable resource supply, use, and management in line with UN (United Nations) Sustainability Goal 12, Responsible Consumption and Production [44]. This shifted priorities away from the initial focus on net-zero, laid out by the funding body, to address environmental sustainability in a more holistic way. All of the above were ingredients in the development of *Circular Chicken* proposition.

Abstracting typical poultry farm operations into a system of resource flows, shown in Figure 6, the *Circular Chicken* poultry house was reframed as a container for inputs and outputs from the perspective of material flows, such as water, to nutrient and chemical flows, such as ammonia. The proposition explored ways to valorise poultry litter, taking ammonia and waste heat from poultry litter for aquaponic production of high value crops. It looked at an existing waste stream from anerobic digestion, a liquid called digestate, and how this might be used to produce micro-algae to feed insects and localise poultry feed or make biomethane for the poultry integrators fleet. It also looked at ways to make mycelium packaging from mushrooms grown in poultry litter and how urban forestry might be used to localise bedding supply. This proposition not only revealed the economic value of existing waste streams but also expanded the houses' boundaries, reconnecting the scale and impact of resource use and waste with global ecologies.

**Figure 6.** Diagram exploring potential material flows to reduce waste and create new economic opportunities. Research for this diagram was funded separately by CIEL to Moy Park.

Reflecting on the existing linear model of poultry production, the design-research team applied circular economy principles to reimagine the future poultry farm as a system of closed loops. The proposal was first described through abstract flow diagrams and sketches. Then through collage images, shown in Figures 7 and 8, visualising the interventions required to facilitate waste-free on-farm production, such as an offset forest and micro combined heat and power plant (CHP). Discussed together, they highlighted the value of designing whole-system interventions that go beyond the redesign of spaces or the installation of technologies. The visualisations opened conversations around the scale of these interventions, whether it is possible to, for example, have an anerobic digestor on each farm or whether one needs to be available to multiple farms in an area to make this option more economically viable.

Viewing productive businesses such as this one as a global resource manipulator, the proposition also opened questions about how Moy Park should respond to current and future climate legislation around carbon accounting, and how the design of poultry housing might help the farm, rather than the business estate, to become an accounting boundary, halting the flow of waste streams beyond this boundary and optimising flows to create new spin-off income streams, such as through the production of mushrooms or algae while targeting existing issues of environmental pollution, such as eutrophication.

**Figure 7.** Sketch section exploring relationships between potential architectural, infrastructural, operational, and technical components on a future "circular" poultry farm.

**Figure 8.** Elevational collage depicting what a "circular" farm might look like and contain.

#### *4.2. Restructuring Actor Hierarchy*

Spatial analysis of the evolution of Moy Park's housing revealed that bird welfare is addressed incrementally within the existing large-scale, tech-centric housing model to ensure high levels of biosecurity. Working closely with poultry behaviour experts, the design-research team developed the *Happy Chicken* proposition, shifting priorities towards welfare and climate resilience, two of the five themes in Section 3.2.

The team played on common architectural practice approaches by developing a client brief for a chicken, which functioned as a tool to give a voice to a silent, non-human actor in the productive system. This playful approach to engagement and brief-making temporarily re-orbited dynamics away from business priorities, providing a flattened hierarchy to expand conversations beyond the current state-of-play. This approach set a mission for the design-research team, animal behaviour, and welfare experts as well as Moy Park and JF McKenna to come together to understand and design for a chicken's dietary, lifestyle, and environmental preferences. Numerous discussions with experts and farmers revealed that, like humans, chickens have different personalities, for example, some are more active and playful than others. Spatial analysis of several types of internal, external, and hybrid housing systems enabled further conversation around designing for chicken happiness. This research was visualised through a series of chicken profiles detailing, for example, how chickens play, what they eat, and daily routines, shown in Figures 9 and 10.

**Figure 9.** Sketches developed with poultry welfare experts to facilitate conversations around internal poultry house design ideas.


**Figure 10.** Sample bird persona developed with poultry welfare experts to facilitate conversations around bird behaviours and preferences across age and breed.

Illustrated in Figure 11, initial ideas focused on creating an integrated forest housing system to resituate more resilient, slower-growing chicken breeds in their indigenous environment, support the immediate offsetting of carbon emissions, and diversify farm income through short coppice crop production. Inspired by organic practices, a time-based design strategy imagined that the future poultry house could be situated within a new forest of oak standards, infilled with hazel coppice. These houses would be constructed from local timber and designed to move through the forest at the end of each production cycle. Flexible netting pulled between trees would allow the chickens to be in a protected outside space within nature. The movement of houses across the forest would be timed to coincide with coppicing and sequenced to fertilise the ground appropriately, while at the same time managing the risk of contamination between flocks.

Retaining the chicken as a key actor allowed the design-research team to question existing processes and scales of production. The proposal opened debate about how designing for bird happiness, through nature-based solutions, may incur higher economic costs. The proposition highlights the biosecurity risk of allowing chickens to live outdoors and seeks to offset this with smaller flocks of more resilient breeds. At the same time, the proposition presented opportunities to diversify income, improve supply security, and support recognisable consumer branding as well as rapid carbon descent through reforestation. Attaching a mission to the design strategy through discussion-based factfinding and profile-building proved effective in helping the poultry integrator to remain open to seeing the value of solutions that were different from current modes of production. Shaking up actor hierarchies opened opportunities to see Moy Park's operations from

a previously unconsidered perspective, even if this approach was deemed economically unviable for current scales of production.

**Figure 11.** Early sketch proposal for a moveable house in a forested environment that repositions the chicken as the primary client.

#### *4.3. Hacking Stakeholder Networks*

Researching U.K. supermarket targets, Moy Park's key customer, revealed plans to increase the number of sustainable, higher welfare meat products as well as plant-based products in stores. For example, Tesco aims to source all soy from deforestation-free regions by 2025, and in the same period, plans to increase sales of plant-based alternatives by 300% [45]. Similarly, research by LIDL suggests that 71% of their shoppers want retailers to be clearer about how the chicken they purchase was raised and 68% of 16–24-year-olds are drawn to plant-based diets [46]. Unpacking these targets created a framework to understand emerging and future expectations around poultry production. Stakeholder mapping, described in Section 3.1.2, supported this notion of the supermarket as the frontier for market access. It also highlighted the complex economic relationship between the integrator and individual farmer, and how this could support or hinder rapid business transformation.

Parallel conversations across the study focused on the importance of maintaining biosecurity and the integrator expressed interest in maintaining a centralised, indoor housing environment to address this concern. Reflecting on the changing consumer-food dynamic, the design-research team sought to reframe this "red line" to imagine a smaller, distributed model for poultry farming, based on creating outbreak exclusion zones to reduce contamination risks and smaller flocks to reduce economic risks if disease outbreaks occur. The *Network Chicken* proposition, visualised in Figure 12 emerged from this concept of designing small, seen, and secure housing. Challenging current centralised models, the concept supports better connections between consumers and their food's provenance. Suitable for urban, domestic sites, the consumer becomes the farmer and custodian of the flock. Managed through artificial intelligence (AI), fed with domestic food waste, monitored by roaming "expert" farmers and visited by a mobile abattoir; the house, or coop, shrinks the gap between consumers and food to enable transparency and visibility in the food-chain. This proposition addresses environmental sustainability through the use of food waste, reducing the need to source feed globally. The proposition raises questions around how this model could be managed, who would own the chickens, and how much would they cost.

**Figure 12.** Early concept sketch proposal for a distributed network of small houses across N.I, monitored and managed by consumers using app technology.

Rather than think of this proposal as an artefact, the design-research team combined current and future priorities of several stakeholders, such as Moy Park's desire for biosecurity and the consumer's desire for transparency. Using icon-style diagrams such as those shown in Figure 13, the team communicated a mix of external influences potentially impacting poultry production. Amplifying these priorities, the *Network Chicken* proposition is extreme and tied to significant cultural and technological leaps. Despite this, it served well to help the wider research team to reflect on how internal and external stakeholders might shape the future of food production and consumption, develop an awareness of the type of emerging technologies likely to catalyse shifts, and explore how current business operations help or hinder rapid changes.

**Figure 13.** Icon diagrams produced as a mnemonic to catalogue and communicate shifts in consumer and technology trends. Top row, left to right: blockchain, weather sensing, drones, off-site management, Internet of things, robotics, alternative protein, RFID-tagging, welfare sound sensing.

#### **5. Discussion and Conclusions**

#### *5.1. Reflections on the Design Methods Used*

Across the project, the design-research team used systems thinking to find pathways from problems to solutions to address the challenge of designing a more environmentally sustainable, higher welfare future poultry house, recognising the "wicked" and complex nature of the challenge. Reflecting on established design methods, the team structured the design process into three iterative, interlinked steps which, starting with baseline analysis of farm inputs, outputs, actors, and networks, then developed themes and scenarios and finally a compendium of solutions. Within each step, a range of design methods, including STEEP, REAP, and the COCD box, were deployed to develop and test ideas. An "Ideation Hourglass" framework supported the development of top-down and bottom-up solutions. Spanning the long and short-term, this framework bridged complementary and conflicting concerns to gather a range of ideas from across the interdisciplinary design team. Different types of visualisations were used as devices to communicate and critique ideas in project team meetings. Diagrams were particularly useful in communicating complex challenges and ideas in an accessible format. Visualisations also flexibly communicated solutions for both near and far-future queries.

Applying different lenses of investigation to the house led to a variety of scenarios and three design propositions exploring the future of poultry housing. Named *Circular Chicken, Happy Chicken,* and *Network Chicken*, each proposed radically different futures arriving from an exploration of different themes and scenarios and reflections on the baseline findings. Each proposition also tested different design methods. The *Circular Chicken* proposition emerged by shifting the focus away from the design of space to the design of flows. Reflecting on the baseline carbon assessment of housing, which found that the carbon embodied in a house with a 30-year lifespan is much less than the carbon to heat and power housing annually, this design proposed integration between house, farm, and estate to enable resource efficiencies through circular economy principles. The *Happy Chicken* proposition restructured actor hierarchies in the poultry production system to design the poultry house for its primary occupants, and then speculated on ways to make this economically and operationally viable. The final proposition, *Network Chicken*, is a reflection of the stakeholder mapping carried out in the baseline analysis. It explores how external actors and technologies might enable decentralised models of production that are bio-secure and better connected to consumers.

Rather than attempt to realistically determine one future poultry house, the propositions tested design methods to explore the future of specific, emerging trends already impacting the poultry production sector. Analysing the poultry house by what flows through it, how it connects to actors-networks and could be impacted by societal shifts revealed that a poultry house is not just a piece of real estate; it is also a home for chickens, a consumer of biomass, a producer of protein, and a node along a supermarket supply chain. By zooming in on the poultry house through different lens, the design approaches taken offered a route to recognise the complexity and interdependency of the challenge yet quickly generate spatial solutions to support changed mindsets that start to break away from locked-in systems. This approach revealed the difficulty of the challenge and that there is no right or clear answer for how the future poultry house should be designed, particularly in the context of climate change and consumer shifts. Though not explored in the scope of this paper, Moy Park carried out sensitivity testing and an internal economic assessment at the end of the project to inform investment strategies. Some of the ideas arising from this project informed contingency planning and strategic visioning while others spurred further in-house research activities on specific environmental sustainability and welfare challenges.

#### *5.2. Designing in a Changing Context*

Undertaking this study through the COVID-19 pandemic and during the Brexit transition period highlighted the speculative nature of planning for the future within an uncertain

present. Global events such as this revealed how quickly some trends accelerate while others dissipate or change. Throughout the project, the concerns of climate change and the vulnerability of the sector to global supply shocks became immediate and apparent to all partners, particularly the poultry integrator. These events included Northern Ireland recording its hottest day on record during the summer of 2021, a shortage of key workers and drivers, reoccurring AI outbreaks, increasing energy prices and a shortage of CO2, all critical to the poultry sector. In addition, the UN's 26th Climate Change Conference highlighted rapidly changing global and local legislation shifts towards environmentally sustainable, high-welfare food production [34,35]. These pressures exposed the fragility of current practices and processes, highlighting the reliance on linear supply chains and tangential industries that are politically, economically, and environmentally volatile.

#### *5.3. Fostering Interdisciplinary Collaboration*

While the research did reveal some interesting ideas and trajectories for poultry production and consumption, one clear benefit of the design methods used was that it supported industry partners to form new perspectives about their sector's future, though aided somewhat by the timing of external events during the project. The design approaches taken helped Moy Park and JF McKenna crash-test pathways to change, shifting from incremental to rapid, radical shifts and identify ways to make "more good" in how Moy Park's operates, rather than tweaking existing systems to make things "less bad" [33]. In a sector that is continually in production, time and effort tends to be spent dealing with problems in the immediate term. This project provided industry partners with the space to consider the day to day in the context of bigger future challenges.

The design-research team tested ways to foster collaborative ideation with industry partners during team meetings. For example, at the beginning of the project, the team started by explored far-future what-ifs rather than unpacking the existing architecture. This approach dismantled the wider team hierarchy bringing everyone together to see problems associated with current production practices by first thinking of big-picture solutions. Discussions remained intentionally informal and playful, often peppered with silly questions challenging the status quo to invite collaborative reflection and build consensus. Together with the methods, this design environment enabled industry partners to "see the forest for the trees", to play out different scenarios without risk, co-create better models based on different value systems, and to map pathways to implement these in future operations.

#### *5.4. Limitations to the Research-by-Design Methods Used*

The spectrum of opportunities that design can unlock—the possible versus the probable—also presents a limitation. The method, shaped by questions of "what" or "who" is placed at the centre of the challenge and is fundamentally determined by the designer's ethical and theoretical position. For example, placing the chicken at the centre of the brief produces a radically different solution to one that places a technology at the centre of the challenge. It is not to say that one approach is better than the other, but that the full spectrum of opportunities can never fully be realised because of the bias that underpins the approach of those involved in the design process. Furthermore, the design methods used throughout this research work require time and space to explore around and beyond the immediate question at hand. In industry, limited time and resources mean that this is not always possible. As such, ensuring mechanisms for exploring the capabilities offered by analysing long-term opportunities is vital in industry and academia. Equally, industry partners must be open and transparent with information and actively engage in the ideation process. Naturally, this openness can lead to tough questions and even tougher solutions.

#### *5.5. Future Work*

This paper highlights the value of research-by-design to reframe, reimagine, and visualise solutions addressing interdependent problems within the poultry sector. The design-research team used an explorative approach, testing a mix of design methods and visualisation techniques within a framework that spanned different scales and timeframes. Informal, playful discussions within the project team fostered ideation between academic and industry partners, bridging top-down and bottom-up action. Further work could expand on the baseline analysis findings or explore further propositional ideas about the poultry house, based on new combinations of themes and scenarios. More work is also required to understand how the design approaches explored in the Results, such as dismantling spaces to flows, might be optimised or compared against one another. Finally, this paper also highlights interesting insights into the value of design-led collaboration between academic and industry partners to consider the future of complex systems.

**Author Contributions:** Conceptualization, E.C., G.K., S.C., A.R., S.B., U.L., B.M. and S.L.; Methodology, E.C., G.K. and S.C.; Validation, E.C., G.K., S.C., A.R., S.B., U.L., B.M. and S.L.; Formal analysis, E.C. and S.B.; Investigation, E.C. and S.C.; Resources, A.R., S.B., U.L., B.M. and S.L.; Data curation, E.C., G.K., S.C. and S.B.; Writing—original draft, E.C., G.K. and S.C.; Writing—review & editing, E.C., G.K., S.C., A.R., S.B., U.L., B.M. and S.L.; Visualization, E.C. and S.C.; Supervision, G.K., S.C., A.R., U.L. and B.M.; Project administration, G.K., S.C., A.R., S.B., U.L., B.M. and S.L.; Funding acquisition, G.K., S.C., A.R., S.B., U.L. and B.M. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Innovate UK, grant number 50015.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** The authors would like to express sincere thanks to Niamh O'Connell for their invaluable knowledge and insight on bird welfare. Complementary research on poultry litter valorisation carried out by the design-research team alongside this project was funded by CIEL. The authors recognize Theano Stoikidou for their input in this separate study which fed into the design process.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

**Eliab Opiyo 1, Santosh Jagtap <sup>2</sup> and Sonal Keshwani 3,\***


**Abstract:** Product design is a key aspect of human intelligence and creativity, attracting not only experts but also people without any formal design training. Although numerous people in developing countries design and manufacture products in metalworking microenterprises in the informal sector, there is still little knowledge about their design process. This paper aims to fill this gap in design knowledge. We aim to investigate the design processes in metalworking microenterprises in the informal sector of Tanzania. In particular, we aim to explore how these microenterprises identify consumer needs and requirements, how they determine the specifications for the product, how they generate and evaluate alternative product concepts, and how they define product details. To address these aims, semistructured interviews were carried out in metalworking microenterprises operating in the informal sector of Tanzania. The findings reveal many facets of their design processes, providing a sound basis upon which design methods and tools can be developed to support their design activities.

**Keywords:** conceptual design; design practice; informal sector; microenterprises; sustainable development; developing countries

#### **1. Introduction**

People around the world are increasingly living in artificial designed environments proliferated with a wide range of designed products, such as electronic gadgets and vehicles, as well as relatively simple products such as kitchen utensils, bicycles, furniture, etc. Design activities and their outcomes profoundly affect our lives and well-being, while impacting social and environmental sustainability, plus the economic growth of organisations involved in design, manufacturing, and other life cycle phases of products [1]. Humans have designed many kinds of products in the past; they are doing this presently and will engage in design activities in foreseeable future as well. Whilst people with formal design education might have more abilities to design products, those lacking formal education also have abilities to design a broad range of products [2–4].

Design processes offer opportunities to design products that satisfy intended requirements while minimising unintended consequences [5]. As such, the scientific field of "design research" not only aims to gain an in-depth understanding of design processes, but also aims at their improvement [1]. As design is a context-sensitive activity [2,6], there is a critical need to investigate design processes in a broad range of contexts representing various social, economic, and cultural characteristics.

However, extant design research has predominantly been undertaken in some specific contexts, typically those found in developed countries and relatively wealthy regions of the world. This has generated limited design knowledge in a narrow range of contexts, while restricting our understanding of design phenomena in the resource-constrained

**Citation:** Opiyo, E.; Jagtap, S.; Keshwani, S. Conceptual Design in Informal Metalworking Microenterprises of Tanzania. *Sustainability* **2023**, *15*, 986. https:// doi.org/10.3390/su15020986

Academic Editor: Yoshiki Shimomura

Received: 16 November 2022 Revised: 21 December 2022 Accepted: 26 December 2022 Published: 5 January 2023

**Copyright:** © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

settings observed in developing countries [2]. This biased contextual focus of extant design research is not just unhealthy for the growth of design research, but also hinders practitioner ability to learn from design phenomena in various contexts and to design products and services for addressing various global challenges. It is therefore crucial to investigate design activities in developing countries, including design activities in informal metalworking microenterprises [2].

Because there are many socioeconomic and cultural differences between organisations in Western countries and those of the developing world, there are large disparities in their approaches to design and how they design products [2]. Knowledge about design activities in informal metalworking firms in the developing world is needed for developing tailored training programmes and methods to support them in designing more successful and sustainable products.

Enterprises in the informal sector provide income-generation opportunities to nearly one-third of the nonagricultural global workforce [7]. Previous research in this field suggests that microenterprises make up a major part of the informal sector. In general, an organization is considered to be a microenterprise if it has no more than 10 employees, and if the majority of its employee are engaged in the manufacturing of products [8,9].

Design research, presently, is conducted largely in developed countries [10], in organizations that are highly dissimilar to informal metalworking firms in the developing world [8,11]. Thus, there is lack of systematic knowledge about design activities in such metalworking microenterprises. This research aims to address this gap in design knowledge. We address this overall aim by attempting to answer the following two broad research questions, namely (a) how do informal metalworking firms in developing countries elicit client needs and identify requirements for products? (b) How do these microenterprises develop design concepts and define detail designs, including materials, dimensions, etc.? To answer these two research questions, we conducted interviews in 24 microenterprises in the informal sector in some selected regions of Tanzania.

Following the above introduction, the remaining paper is organised as follows. Section 2 reviews the related literature, highlights the need to investigate design activities in a broad range of socioeconomic contexts, discusses the role of the informal sector in developing countries, and identifies the crucial need to explore design activities in informal metalworking microenterprises in developing counties. Section 3 provides the details of the research method, presenting information on sampling, data collection, and data analysis. Sections 4 and 5 present the findings of a qualitative analysis of interview data collected from 24 informal microenterprises. Finally, Section 6 discusses the findings, along with concluding remarks and limitations of this study.

#### **2. Background Literature**

This section presents the literature review on the design process and design research, specifically in the context of product design in informal sectors. This review was conducted using a semisystematic approach [12]. An author of this work has previously published an integrative literature review in the domain of design and poverty, specifically on role of poor people in formal and informal design sectors [2]. This integrative literature review guided the semisystematic literature review in this work. This ensured that appropriate literature is accurately covered to be able to answer the research questions.

#### *2.1. Design Process and Design Research*

Designed products are recognized as one of many important means for a society to advance its economic and social well-being [1,5]. As individuals, we appreciate a well-designed product, and respond to it by selecting it from available products [1]. The ability of individual entrepreneurs and the industry at large to design efficient, inspirational, and aesthetic products is therefore crucial. The ability to design products is recognised as one of several forms of creativity and a fundamental aspect of human intelligence [3,4]. Consequently, it is important to understand what people do when they exercise design ability [13].

In design research, in order to fully understand how people actually design products, empirical data on the design process are collected by using a variety of methods, including interviews [3,10]. Some characteristics of design processes have been widely observed. It is commonly accepted that the design process is iterative [14,15] and that identification of consumer needs, the generation and evaluation of concepts, and defining of product details are important activities in a design process [2,14,16].

A large body of research has demonstrated that the design process provides a maximum scope to improve, modify, and create new products because many crucial decisions are made in this very process [3,5,13]. Deeper scientific knowledge on the design processes is needed to develop design knowledge [1,13]. Several authors argue that generating design knowledge requires investigating design processes in a variety of settings, covering a broad range of sociocultural and economic environments in which the products are designed [17,18]. However, despite the prevalence of design activity, extant design research is predominantly focused on individuals and enterprises in developed countries, thus investigating design processes in some specific socioeconomic and cultural settings [8,10]. This narrow focus of design research has constrained the generation of design knowledge by limiting its focus [8,10]. Addressing these gaps in the existing design knowledge necessitates that design research be undertaken in a range of fields, including, among others, informal sector microenterprises in developing countries. Such microenterprises are distinctly different on many dimensions from enterprises in developed countries.

#### *2.2. Informal Sector*

Low-income people in developing countries may produce a variety of products, such as furniture, utensils, and common domestic goods. These activities of making products create jobs, supporting them to satisfy their unmet or under-served needs [2].

In developing countries, people generally work either in the informal or formal sectors. A person working in the formal sector has a formal contract with the firm owner or employer, typically receives a decent monthly salary, and generally has access to a social security system [2]. In contrast, a person working in the informal sector typically lacks legal contract with the firm owner or employer, lacks a well-ordered work environment, has an irregular work duration, typically receives a low and irregular income, and has no access to social security [2,19].

As mentioned earlier, researchers from sociology, anthropology, and economics, as well as from other disciplines, have extensively studied the informal sector [18,20]. Furthermore, in recent years, studies have been carried out on firms operating in the informal sector [19,21,22]. However, research devoted to this subject is relatively minor as there is widespread commonness of informal firms in developing countries. Recent studies suggest that the contribution of the informal sector to the GDP in developing countries is about 40–60% [23]. The informal sector employs just under one-third (31.5%) of the nonagricultural workforce around the world [2]. As such, the informal sector represents a large segment of the total global economy. For some authors, informality is a "voluntary" choice [24], whereas, for others, informality is an "exclusion" [25].

Some scholars have examined the informal sector as a labour phenomenon [22,26], examining its role in creating jobs for numerous low-income people in developing countries, as well as its failure in providing access to various benefits, including social security. Other scholars have dealt with the challenge of how to describe and define it, its relationship with other parts of the economy, and how it affects growth [7,27]. In addition, some researchers have studied the informal sector as microenterprises, which can benefit from the provision of technology and funding [28]. These studies suggest that microenterprises make up a significant fraction of the informal economy, and most of these firms manufacture products [8,9]. To sum up, the informal sector is a significant part of the total economy in developing countries. However, design researchers have given little or no attention to this sector, despite the presence of product design activities in this sector [2].

#### *2.3. Product Design in Informal Microenterprises*

Researchers generally agree on the main traits of manufacturing microenterprises operating in the informal sector. They typically use readily available tools and equipment, are labour-intensive, depend on family ownership, rely on competitive and unregulated markets, and are small-scale [2,29]. They face many constraints, such as lack of access to financial and material resources, weak infrastructure, weak marketing and organisational resources, and absence of specialised knowledge and skills [8,29]. Despite these constraints, studies have suggested that informal microenterprises have the ability to design products [8,9,30–34].

Despite the acceptance of design ability in microenterprises in the informal sector, there is little or no research on their design activities. This can partly be clarified by the perspectives of the studies undertaken in the informal sector. These studies are generally undertaken from an economic perspective and consider design activities of the microenterprises as a "black box" [2,8,9,34]. A recent study has identified the need to investigate design activities in informal microenterprises, including metalworking informal microenterprises [2].

#### **3. Method**

In general, design research studies use the following research methods: (a) design experiments; (b) questionnaire analyses (online and offline); (c) interviews (online, physical, or telephone); (d) diary studies; and (e) observational studies in real settings.

In this work, we have used interviews as a research method to collect data. These interviews were conducted in a real setting. This allowed us to obtain the first-hand experience of the facilities available with these firms. Additionally, we saw the products designed by them and had access to their documents, such as sketches, photos, and design drawings. This also allowed us to dynamically react to the responses and further probe, if required, through open-ended questions—something not possible in the design experiments and questionnaire analyses that are usually conducted in lab settings. Similarly, conducting diary studies or 43 observational studies in real settings for 24 firms would have been time intensive.

The interview method—which is widely used in design studies [35]—was used in this work to explore how designers elicit customer needs; how they formulate requirements for products; how they generate, evaluate, and select design concepts; and how they specify product details. Interviews took place between March 2021 and January 2022.

Random-sampling technique was adopted and used to select 24 informal metalworking microenterprises in the Coast and Dar es Salaam regions of the United Republic of Tanzania. The microenterprises in which interviews were conducted were randomly selected to avoid biases and to ensure that the eventual findings approximated those of the actual population. All the interviewed microenterprises were informal metalworking microenterprises, with no more than ten permanent employees. Direct communication and chain-referral sampling approaches [35] were used to recruit interview subjects. Table 1 presents the information on the selected microenterprises and the subjects who participated in the interviews. The major job orders that informal metalworking microenterprises receive include supplying building construction resources (e.g., aluminium windows, door frame grills, aluminium door frames, steel gates, doors, movable kiosks, office space partitioning, and window grills), metal furniture (e.g., bed frames, reading table frames, dressing tables, chair/couch frames, and TV stands.), household utilities (e.g., cooking utensils, charcoal cooking stoves, and charcoal grill stove), agricultural equipment (e.g., chicken feeding utensils), and light machinery (e.g., grain shelling machines, flour milling machines, and brick-making machines)—see Table 1. Figure 1 shows examples of products produced in studied microenterprises.


**Table 1.** Information on informal metalworking microenterprises and on interviewed participants.

**Figure 1.** Examples of products produced in studied microenterprises.

Semistructured interviews [36,37] were conducted with two subjects who were familiar with design practices in the firm. One subject was the main speaker and the other was there to corroborate the accounts given by the main speaker and to provide clarifications or any additional information whenever required. As such, a total of 48 subjects participated in interviews. It should be noted that this was qualitative research and the intention was not to gather data or information for statistical analysis. The focus was, rather, on acquiring a proper understanding of the actual design practices through qualitative research. Only 5 of 48 interview participants held bachelor's degrees. All respondents were male, and most of them lacked technical or design training.

Interviews took place at the interviewees' places of work. This allowed the interviewers to informally observe the working practices and culture in informal metalworking microenterprises. The interviewees were asked to refer to particular projects during interviews. Figure 2 shows examples of working practices and conditions in these microenterprises. The mean duration of the actual interviews, excluding briefing and debriefing, was 51 min. We sought consent of the subjects to participate in the study, to audio- and video-record the interviews, to take pictures, and to use gathered data/information in analyses and publications.

**Figure 2.** Some examples of working practices and conditions in informal metalworking microenterprises.

Interviews in all 24 microenterprises were administered by using the Swahili language and were audio-recorded. The recorded audio contents were then transliterated in Swahili and eventually translated into the English language. A general inductive and iterative approach was then used to analyse the translated transcripts [36]. The analysis was content-driven.

Sections 4 and 5 present the results of the analysis. Excerpts from the transcripts of the interviews are included to illustrate the findings. Some of the excerpts of the interviews have been edited for ease of comprehension and any additional information is included in brackets.

#### **4. Identification of Consumer Needs**

There is little knowledge about how informal metalworking microenterprises in developing countries elicit customer needs [38]. In this section, we present findings on how the designers in microenterprises elicit customer needs and how they formulate and organize the requirements—which form the first part of the research questions of this study.

Interviews were conducted in the firms listed in Table 1. Generally, design processes in these informal metalworking microenterprises entailed interacting with clients with a view to gathering their wishes and preferences with respect to functionality, dimensions, material type, aesthetic features, and cost of the product.

#### *4.1. Elicitation of Needs*

We explored how the needs of the customers are identified. Elicitation of needs in these microenterprises essentially entails conversion of tacit and subjective customer verbatim constructs into needs statements. Needs and requirements for products in these firms largely originate from interactions and one-on-one conversations with customers, as evidenced by the following sample interview excerpts.

The client paid visit once to the [our] firm to press order, but we also visited the client to verify the space dimensions [Firm I, movable kiosk] ... I interacted with the client straight away by discussing the picture that he brought and we also paid the visit to the client to take measurements. [Firm M, fence gate]

Some of the requirements originate from existing products. Customers use sketches, engineering drawings, and pictures of existing products to describe how they wish their products to be. Pictures are widely used as stimuli for discussion—refer to the interview excerpts below.

The client came with pictures of an existing bed frame and we discussed and changed its appearance and added seating feature [Firm K, bed frame] ... After seeing the pictures and based on our personal experiences we recognized the needs [Firm L, gates and window grills] ... the client used the picture to explain how she wanted the dressing table to be like [Firm N, dressing table] ... the client brought a picture of window grill with decoration features he wanted". [Firm X, Window grill]

The designers and their clients refer to the pictures during discussions and agree on additional features to incorporate in the final designs.

For products with novel features, the design processes start from scratch. Some microenterprises use few requirements to produce trial products that they showcase to potential customers. Designers and customers use these trial products as stimuli for discussing how the product should be—see quotes below.

For this product we developed a [sample] product and interacted with potential clients mostly during exhibitions [to explore its acceptability]. We visited several exhibitions to showcase the prototype. [Firm C, foldable multipurpose furniture]

Some designers visit clients and interview them at their sites—see sample excerpt below. The number of outgoing and incoming for such visits vary from two to ten. This allows the designers to know the use environment of the product.

The client paid visit once to the (our) firm to press order, but we also visited (the client) to verify the space dimensions. [Firm I, movable kiosk]

The effectiveness of needs elicitation usually depends on the technique adopted and used [39–45]. Informal metalworking microenterprises use unstructured interview methods and appear not to prefer using formal alternative needs elicitation channels such questionnaire surveys, structured interviews, and maintenance reports.

#### *4.2. Interviews with Customers*

Interviews with customer is the approach that is widely used to collect needs data in informal metalworking microenterprises. Some researchers claim that interviews particularly structured—are one of the most effective needs elicitation techniques [39]. Interviews enable the informal metalworking firms to get glimpses into how customers wish the product to function and be—see the representative excerpt below.

We discussed and agreed with the client on how the gate should be like and how different [the client] wanted it to be compared to the pictured one. [Firm Q, aluminium windows]

Generally, it can be said that the needs originated from interviews with customers, which generated first-hand data that are used in the formulation of requirements. During conversations with clients, firms typically record first-hand needs data by taking notes, sketching concepts, or taking photos, as substantiated by the sample quotation below:

We took notes [... ] typically hand-written text complemented with hand sketches to describe concepts. [Firm A, palm oil filter]

However, as demonstrated by the excerpts below, some firms do not document conversations with clients and only just listen and recollect what was said later on.

...... . we just heard and recalled what the clients had to say. [Firm J, chicken feeding utensil] ... We didn't sketch anything, ... we just got word-of-mouth explanations on the needs. [Firm G, Charcoal stoves, barbeque ovens, water gutters, metal suitcases]

Generally, the designers typically receive verbal explanations from clients but do not archive the conversations. They regularly use hand-written text, pictures, or hand-prepared sketches to describe requirements—see the representative excerpts below.

[Customer came] with the sample photo or sketch. [Firm B, peanut peeling machine] ... past experiences ... helped us to identify requirements, and we then prepared sketches of the gate. [Firm T, fence gate]

Traditionally, in order to conduct meaningful needs analysis, it is imperative to properly document conversations. Documentation can be in the form of textual interview transcripts complemented with hand-sketches to record nonverbal information [42].

#### *4.3. Sources of Requirements*

In informal metalworking microenterprises, requirements originate from different sources. Apart from interviews, standards and use of past experiences are also the sources of requirements in many microenterprises. Some requirements also originate from design constraints and from the designer's own expert knowledge. Reusing past knowledge and experiences seems to be a natural way to elicit requirements and of handling uncertainties in informal metalworking microenterprises. Moreover, the requirements gathered in informal metalworking microenterprises describe technological, social, environmental, and economic aspects of the product.

#### *4.4. Prioritization of Requirements*

Designers in informal metalworking microenterprises identify, at most, eight requirements. The identified requirements describe functionality, appearance, dimensions, and

cost of the product, but are generally neither properly worded and documented, nor archived. The requirement statements do not properly describe the product.

Customers are involved in prioritizing the requirements. Apparently, the requirements that describe dimensions, appearance, cost, and functionality of products are given higher priority—see some of the representative excerpts below.

The appearance and dimensional specifications were given higher priority [Firm N, dressing table]. The strength of materials and dimensional requirements were given higher priority. [Firm W, meat grilling oven]

However, some microenterprises do not prioritise requirements, but give equal priority to all identified requirements.

All requirements were given equal priorities. [Firm P, windows grill]

In engineering design, requirements are typically prioritized systematically by using formal methods such as Quality Function Deployment (QFD) and Analytic Hierarchy Process (AHP) to help designers set definite goals for their products and to ensure that the design project focuses on and addresses the key customer's needs [39,42,43]. Our findings reveal that the informal metalworking microenterprises did not employ such methods.

#### *4.5. Difficulties Faced in Identifying Needs*

Some designers in informal metalworking microenterprises felt that the process of identifying needs is tedious and expensive. Some excerpts from the interviews to explain the difficulty faced are presented below.

Our approach entailed making prototype and demonstrating concept. ... was tedious ... . [expensive], and ... [lengthy]". [Firm C, foldable multipurpose furniture]

Other difficulties faced include the identification of needs that cannot be met due to limitations of capabilities of manufacturing equipment within the microenterprises, and customers disputing some of the obvious requirements, e.g., those based purely on technology-related and economic reasons—see some of the interview quotations below.

The customer was [... ] disputing technical requirements [... ] wanted the product to be painted without spraying red-oxide paint first to cut cost ... . [Firm D, door gate]

It was also observed that designers in informal metalworking microenterprises also confront numerous inherent difficulties typically faced in identifying customer needs and requirements. These include ambiguity and imprecision of requirement statements [40], uncertainty of whether needs are genuinely captured [44], and excessive focus on the technical details of a product [41].

#### **5. Conceptual Design**

In a formal design process, requirement identification is followed by a conceptual design stage. According to Pahl and Beitz [45]'s model of designing, the conceptual design stage involves creating function structures, identifying working principles, and combining the working principles into a working structure. Additionally, conceptual design focusses on novel idea generation.

Researchers converge on the view that *novelty* is a measure of the newness of a product with respect to existing products in the market satisfying the same function [46]. For instance, the first pin-hole camera, the first X-ray machine, and drugs such as penicillin are regarded as very highly novel products [47].

Industries are under huge pressure to launch novel products because of reasons such as increased competition and customer expectations, rapidly changing technology, and shorter product life cycles [48,49]. To generate novel designs, designers use methods such as brainstorming, biologically inspired design, TRIZ, functional analysis, etc. [50]. Apart from these methods, designers take inspiration from sources such as patents, expert opinions, discussions with colleagues, and past experiences [51,52]. Greater product novelty, in turn, positively influences the product quality, which in turn determines the market share of a product [53–56]. Therefore, in formal design methods, novel idea generation is central to conceptual design.

This contrasts with the informal design process in which the decisions are driven largely by intuition and past experiences. In addition to following informal design processes, the microenterprises interviewed in this study functioned within constrained resources and limited technical proficiency. These factors are likely to influence the design outcomes. Overall, this section, describes the findings on how these firms generate and evaluate alternative product concepts, and how they define product details—which form the second part of the research questions of this study. More specifically, Sections 5.1–5.3 present the findings of our investigation with regard to the conceptual stage of the design process followed in the informal metalworking firms in Tanzania. Section 5.4 focuses on methods adopted by these firms for material selection and determination of dimensions.

#### *5.1. General Approach for Concept Generation*

We intended to abstract a general approach to concept generation, which the microenterprises in Tanzania follow. The employees of the firms were asked to brief the steps that they followed after requirement formulation. Sixteen firms responded that they manufactured the product straight away after requirement formulation. Six firms reported generating concept sketches for the purpose of communicating the design to the clients. We present some excerpts from the interview as evidence.

I started to manufacture the sample product straight away after knowing the requirements and dimensional specifications. [Firm F, Aluminum windows, doors, aluminum-frame furniture]

We proceeded to take measurements at the site, calculate costs, purchase raw materials and then we started to manufacture the product. [Firm R, Windows, door grills and wardrobes]

Clearly, this contrasts with the formal conceptual design process, which involves activities such as the exploration of concept space, evaluation of generated concepts, creation of prototypes, and a preliminary analysis of the proof of concept.

#### *5.2. Concept Generation*

We explored how designers in informal metalworking microenterprises generate concepts. Findings on how they explore and select concepts are presented below.

#### 5.2.1. Exploration of Concepts

To understand whether the concept space was explored, we inquired about the number of alternative concepts generated. Twelve firms reported generating one concept and improvising it over multiple iterations after discussing with the clients. Three firms reported generating four concepts. Two firms reported generating six and nine concepts, respectively. Overall, the majority of the firms generated only one concept. Some excerpts from the interview to substantiate this are presented below.

There were no any alternative concepts. The client wanted the kiosk to appear and be produced as the one shown in the picture that was provided. [Firm J, Chicken feeding utensils, charcoal stoves, metal suitcases]

We prepared only one alternative concept and improved by continuously engaging the client" [Firm V, doors and windows grills, gates]

In design creativity research, the number of ideas generated (also known as idea fluency) directly correlates with the quality of ideas [57–60]. Design methods such as brainstorming support idea fluency and are widely used in industries [61]. As most of these firms generated only one concept, it may be unlikely that these concepts would be of high quality.

We intended to understand if the firms had awareness about novelty and innovation. Most firms agreed that innovativeness involves new methods, new designs, the use of new materials, or the use of new technology. It can be inferred that these firms had an understanding that innovativeness involves newness. However, interview results revealed that they did not use scientifically established design methods for novel ideation. They depended on stimuli from social media sites, competitors' products, and discussions with colleagues. Only three firms reported using technical standards, engineering drawing and design methods such as brainstorming. When asked about the use of any special methods to design creative products, they did not report the use of any different methods.

#### 5.2.2. Concept Selection

In a formal design process, designers use the established methods for concept selection. Concept evaluation methods, such as Pugh's method and rank-ordering method, use requirement satisfaction as criteria to evaluate concepts.

In the case of these firms, our observations were the following. Eleven firms reported that the concept was selected based on discussions with the client. Six firms said that concept selection was not required, as they generated only one concept and iterated over it. Two firms seemed to be aware of subfunction concepts and stated that they selected concepts based on discussions with the end users. Here are some related excerpts from the interview.

For various sub-functions, there were alternative concepts, from which best were selected through discussions that involved both potential end users, and the final composition of the entire product concept was eventually generated. [Firm D, Furniture, doors, shoes-stands, window grills]

By involving the client (who had the final say) and based on our past experiences. [Firm X, window and door frames, gate frames, cooking stoves, railings]

Furthermore, we intended to understand which product attributes they deemed as important for concept selection. Nine firms said that functionality was the key attribute for selection of concepts, seven firms reported that aesthetics was the key attribute for concept selection, while four firms reported durability as an important attribute.

#### *5.3. Concept Representation*

The concepts are represented using various modes, such as by using sketches, clay models, paper models, and CAD models. These representations serve various purposes, such as active learning, refinement, communication, and exploration [62]. From the interviews, it appeared that, in the cases of the interviewed firms, the concepts were represented primarily to communicate the design to the client and to take the approval for manufacturing. Concept representation seemed optional, and the ones that did mostly used hand-drawn sketches as the representation medium. Nine firms said that they manufactured the product directly without representing it, seven firms reported using concept sketches, and one firm reported making CAD models. Related excerpts are presented below.

We arrived directly at the final solution—based on past experiences. [Firm T, Door and window grills, bed frames, fence gates]

We directly manufactured the product according to the agreed dimensions and other requirements. [Firm U, Aluminium profiled doors and windows, aluminium furniture, deck rails]

Most firms did not formally document the sketches. Upon request, we collected sketches from three firms only (See Figure 3). It can be seen that the people who prepared most of these sketches lack the skills required for industrial sketching.

**Figure 3.** Examples of concept sketches produced by some of the firms interviewed in this study. (a) Firm D manufactures furniture such as chairs, bed frames, doors, shoe stands, and window grills. (b) Firm E manufactures door and window grills and furniture; employee sketching the concept on floor. (c) Firm I manufactures bed frames, door grills, and door gates.

#### *5.4. Material Selection and Dimension Identification*

With regard to main considerations for material selection, most firms used square hollow sections of steel and aluminium. Material cost, strength, durability, and availability were found to be the main considerations for material selection. The choice of material was made by the client. With regard to calculation of dimensions, some dimensions were established by visiting the site, while the other dimensions were determined based on past experiences. One firm also reported that they arrived at the dimensions by studying the competitor's product.

In a formal design process, material cost, strength, and availability are some of the governing criteria for material selection. Scientific methods and tools of material selection, such as the Cambridge Engineering Selector (CES) toolkit are widely used for material selection. None of the firms reported using such methods. Similarly, for the calculation of product dimensions, designers used methods such as quality function deployment [62], and referred to anthropometric data [63], industrial standards, and government regulations. However, in our study, only one firm reported using ergonomic standards.

#### **6. Discussion**

#### *6.1. Needs Identification*

Traditionally, needs elicitation in large formal enterprises is a well-organized process [64–66], which passes through definition of scope, raw data gathering, derivation of needs statements, and arrangement of needs according to their importance.

We explored how designers in informal metalworking microenterprises elicit consumer needs and requirements for products. We found that designers in these firms mainly use interview methods to identify needs and requirements for products. Overall, the designers only use their intuitions to identify the needs and are generally neither unaware of the existence of a formal process for eliciting needs nor of the tools and methods of identification of needs and requirements. Furthermore, we found that needs are not systematically interpreted and translated into requirements. Specifications are not quantified through a formal process that traditionally entails identifying metrics and measurement units, which corresponds with the needs or requirements. Specifications are not necessarily tied to the needs and requirements. The dimensional specifications for the products in these firms are established based on past experiences or by taking actual measurements from the sites. Other specifications are also determined in an ad hoc fashion.

Product specifications are, in fact, the product attributes or design features [67]. Formulation of specifications for a product is one of the mainstream early-stage design tasks, through which concrete specifications are determined based on customer needs [68]. Quality function deployment (QFD) [69] is commonly used to determine the specifications for the products based on the needs. QFD utilizes house of quality (HoQ) to map customer needs and requirements to the specifications [70]. Other methods used include semantics methods [71,72], which apply.

Other identified shortcomings in needs elicitation include (1) the absence of mission statements; (2) assumptions that constrain the development efforts; (3) excessive reliance on the one-on-one interview method, which is known to have inherent drawbacks such as biases and reliance on interviewer capability [33,37]; (4) the absence of needs data that describe sensory experiences, such as comfort or style; and (5) ambiguity in the needs statements.

#### *6.2. Conceptual Design*

We explored how informal microenterprises generate and evaluate alternative design concepts and define product details. We found that the number of concepts generated by informal metalworking microenterprises were limited. In most cases, only one concept was generated, which was iterated over multiple cycles of improvement. We propose the following reasons for the generation of a limited number of concepts.


Designers use various concept-representation modes, such as paper models, cardboard models, sketches—digital and physical, clay models, and CAD models [62]. For these modes of representation, concept sketches seemed to be the dominant method for concept representation in the case of the interviewed microenterprises. These sketches were informally documented with people who seemingly lacked the skills needed for industrial design sketching. Overall, the limited use of concept-representation methods might be due to lack of proper training and required resources. Further, it might be that the designers had prior experience with the manufactured products, so they did not feel the need to represent these designs in a detailed manner.

While scientifically established design methods such as SCAMPER and brainstorming were not used, the major sources of stimuli that they reported were social media sites, discussions with colleagues, and competitors' products. Again, this indicates a lack of training, awareness, and required skills for a designer.

Functionality and aesthetics seemed to be important product attributes for them. They determined critical dimensions by taking measurements from the use environments of the envisioned products, by using their own past experiences, or by adopting dimensions of the existing products. Overall, materials are selected by the designers, but customers have the final say. The customer also has the final say in deciding on features of the product. Additionally, those surveyed appeared to be familiar with the concepts of novelty and innovation.

Our results corroborate earlier research, such as in [8,29], which reported that informal sectors avoid exploratory activities such as prototyping or tinkering due to costs and available resources. Additionally, researchers have emphasized on the need of co-designing at every phase of design process, especially for sustainable impact on marginalized societies [73]. In our study, we observed that these firms maintained regular interactions with their customers, who seemed to have the final say in the design and manufacturing of the product. This points towards the existence of co-designing practices in the informal sector.

#### **7. Conclusions**

This research investigated the design process of the informal microenterprises in Tanzania—in particular, the need identification, requirement formulation, and conceptual design phases of a design process. For this, we interviewed 24 metalworking enterprises at their workplaces in real settings. With regard to need identification and requirement formulation, it was found that these firms follow their intuitions to identify the needs and are generally neither unaware of the existence of a formal process for eliciting needs nor of the tools and methods of identification of needs and requirements. With regard to conceptual design, it was found that concept exploration and ideation methods are not followed. In most cases, they receive routine job orders, for which ideation in not considered to be as important. For both, need identification and conceptual design, these firms rely heavily on the past experiences and in general lack the necessary tools and training.

Additionally, this research has revealed that design activities, e.g., activities associated with needs identification and conceptual design, are influenced by the context in which those activities are performed. For instance, the informal metalworking microenterprises studied in this research face various constraints, such as a lack of formal design education and weak access to various resources. These sociocultural and economic aspects of the context have an influence on their design activities.

A limitation of this work is that the questionnaire used was written in English and translated into the Swahili language. The transcriptions of the responses in the Swahili language were translated into English. It is probable that these translations might have altered the meaning of some words.

This research raises several interesting questions for further research, some of which are discussed below.


of English. This calls for the necessary development of smartphone-based design methods and training programmes in local languages.

Overall, this research paves the way for a number of future research avenues in terms of developing dedicated design models and methods for constrained environments, studying tacit co-design and development of low-cost design tools in local languages. Also, it is likely that the results from this study hold true for resource-constrained enterprises in other developing countries.

**Author Contributions:** Conceptualization, E.O,. S.J. and S.K.; Methodology, S.J. and S.K.; Formal analysis, E.O. and S.K.; Investigation, E.O., S.J. and S.K.; Resources, S.J.; Data curation, E.O.; Writing original draft, E.O., S.J. and S.K.; Writing—review & editing, E.O., S.J. and S.K.; Supervision, E.O., S.J. and S.K.; Project administration, S.J.; Funding acquisition, S.J. All the authors have contributed equally to this research. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Swedish Research Council grant number [2020-03353] And The APC was funded by Swedish Research Council.

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** Data supporting reported results can be found on request from the Department of Mechanical Engineering, St. Joseph University In Tanzania, Dar es Salaam, Tanzania via ezopiyo@gmail.com and some of publicly archived datasets can be accessed via https://drive.google.com/drive/u/0/folders/1mh1PfHejEwWzMm3\_POGksdc8FXxbEunD (accessed on 15 November 2022).

**Acknowledgments:** This work was funded by the Swedish Research Council.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

#### **References**


**Disclaimer/Publisher's Note:** The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
