Design for Six Sigma in the Product Development Process Under a Sustainability Point of View: A Real-Life Case Study

Abstract

The modern economic landscape, now more competitive than ever, compels companies to create increasingly innovative, cutting-edge, and cost-effective products. In this regard, the design and development phases play a crucial role, as they closely determine the final satisfaction of the customers. It follows from this the need for a structured approach to product development, which allows companies to identify product key characteristics, also useful when there is the need for carrying out the redesign of an existing product. This work provides the application of the Design For Six Sigma (DFSS) methodology to a real redesign case study, based on the improvement in a specific mountain bike model (baseline product). The final target is identifying the main criticalities and intervention areas for the bicycle, to provide valuable suggestions to designers and developers for creating a new product alternative characterised by an extended horizon to various markets. DFSS is applied to identify and optimise Critical-To-Quality (CTQ) features, aiming at making the project as responsive as possible to customers’ needs. More specifically, two main tasks are addressed. The first one is the application of the identify phase of the IDOV approach (Identify, Design, Optimise, Validate) to identify the most pertinent Voice of the Customers (VOCs) to be implemented in the product. The second main task is the analysis of the integrated product development process carried out through the Six Sigma (SS) methodology, to transform the VOCs into CTQs via the Quality Function Deployment 1 (QFD1) and to prioritise the CTQ to achieve design excellence. The identified improvement strategies are presented and critically discussed in relation to their potential to reduce the environmental impact of the overall Life Cycle (LC) of the product.

1. Introduction

The economic success of a product is strictly related to the quality of thinking that has generated it [1]. More specifically, it strongly depends on the ability of the production company to identify the key features of the product itself and to translate them into design parameters able to satisfy the customer needs in the most convenient way from an economical point of view [2,3]. At the same time, the development of a product is the result of a carefully studied process, which encompasses the entire set of activities between identification of market opportunities, production, sale, and final delivery [4]. In this context, the use of a structured methodology for product development represents the starting point for continuous improvement. This is confirmed by the fact that many of the most advanced companies in the current Italian economic landscape embrace this approach, namely “the recurring and targeted activity aimed at increasing the overall performance of the system” [5].

The literature provides various methods for the identification of design parameters for customers’ satisfaction in different contexts, such as virtual video, fast method, fuzzy approach, and many others [6,7]. In this context, the traditional product development process can be summarised in a series of six steps. The first one is product planning, often referred to as phase zero, it leads to the production of a portfolio of potential projects, and it defines which of these have to be undertaken in the short, medium, or long term. The second step is the conceptual design, which encompasses activities needed to draft a product development plan. The purpose of such a stage is to select a concept (for example, in terms of shape, function, and product features) and combine it with a set of measurable specifications that describe product requirements. Thirdly, the embodiment design, whose goal is the preliminary design, is often called “system-level design”. In this phase, engineering and architectural elaboration of the concept takes place, which must be in harmony with the final specifications and economic analysis. The fourth stage is the detailed design, which consists of the collection of all documents, drawings, and files needed to create the product’s technical dossiers. The analysis continues with testing and refinement. Such activity is necessary for project technical validation, as well as to ensure that the product effectively meets the design requirements, and it encompasses the realisation of prototypes and testing activities. Finally, the production ramp-up. Prior to actual production, it is the step in which the product is made by means of the final process: the objectives are multiple, from training activities for production personnel to addressing issues raised by them.

In the context of product development and improvement, Six Sigma (SS) is a widespread methodology to achieve high-quality production standards and low costs, maximising the “Total Customer Satisfaction” [8,9]. Originating in the mid-1980s, SS focuses on reducing management costs and warranties by shortening production times and minimising defects [10,11,12]. Such advantages are obtained by lowering production variability, improving quality, and reducing non-conformities by means of standard deviation as a measure of process performance [13,14]. Strictly related to SS, Design For Six Sigma (DFSS) represents an evolution of the method, in the sense that its main target is implementing the design phase improvement strategies identified by SS tools [15,16,17]. From a practical point of view, DFSS applies an operating mode, which allows designing for production, thus providing a combination of customer requirements with process capability [18,19]. To this end, DFSS encourages the use of innovative methods throughout the overall design process, with the purpose of limiting the use of a trial-and-error approach [20,21,22]. When considering the implementation of DFSS in specific case studies, the knowledge of all available optimisation methods is a necessary condition, since it is not automatic that they can be properly adapted to all types of applications. Therefore, it is the responsibility of the engineer/designer/manager to identify the appropriate tool for the specific situation, and often the synergistic use of several tools is the right choice for achieving the shared goals [23]. When applied to the development of new products, DFSS consists of the following main phases, which represent the IDOV approach [21]: Identify, Design, Optimise, and Validate. The Identify step involves defining a work team, establishing the business case and drafting the project plan. Subsequently, customer needs are collected and translated into CTQs to identify those critical for achieving the established objectives. Following, the Design stage aims at analysing the CTQs and establishing the Functional Requirements (FRs) (e.g., the high-level functions of the system to be designed). A mapping of FRs into Design Elements (DEs) is performed, generating and selecting the most suitable concept. In the Optimise step, a detailed design is carried out to create a robust design, reducing the effects that variability causes without necessarily eliminating them, thus ensuring more stable and controllable design features. The last phase (Verify) includes project testing and validation: prototypes are tested to ensure the required quality standards and verify that the product meets the project objectives.

One of the key tools of DFSS is the QFD [24,25]. QFD aims at translating the Voice Of the Customers (VOCs) into measurable engineering parameters, establishing their priority, and ensuring communication between departments involved in the production process [26]. This ensures that decisions are guided, guaranteeing that the interests of all stakeholders are properly taken into consideration [27]. The QFD metric consists of a graphical representation shaped like a house, comprising five main elements: -House A, containing the list of previously collected VOCs, which are appropriately prioritised; -House B, concerning the VOCs benchmarking table and providing a concise overview of the strategic market objectives for new products; -House C, encompassing the list of CTQs developed based on the requirements, usually structured in a tree diagram with two or three levels; -House D, a prioritisation matrix aimed at modelling the relationship between VOCs and CTQs: each cell of the matrix represents a judgment expressed on the strength of the relationship between VOCs and CTQs; -House E, including the interaction matrix between CTQs; -House F, holding the CTQs benchmarking table and an elaboration that allows prioritising CTQs. House F also identifies CTQs targets needed to satisfy the customers and appropriately calculates the Customer Satisfaction Index (CSI), which represents the customers’ satisfaction level for the considered solution. To combine VOCs into VOP (Voice of the Process), a four-stage scale is used, where each step takes the output from the previous one as its input. Each input is evaluated through the QFD, which has the task of finding a relationship between input and output [28].

The application of QFD to analyse customer satisfaction reveals the critical role of service quality dimensions in aligning product attributes with consumer expectations [29]. Furthermore, the incorporation of QFD, value engineering, and lean approaches in prioritising control tests for product design exemplifies the confluence of quality engineering and productivity enhancement in the manufacturing domain [30]. In this regard, an example of application is the healthcare sector, where the algebraic operations of QFD house-of-quality are leveraged to prioritise Industry 4.0 technologies integration in hospitals, thereby fostering a systemic digital transformation [31,32]. In the context of technology transfer, the amalgamation of Fuzzy QFD with a Fuzzy Inference System (FIS) offers a large framework for licensor selection, enhancing organisational capabilities and market performance [33,34,35,36,37]. Another application field of QFD-based methods is the development of manufacturing information systems, where several studies underscore the importance of systematic approaches in ensuring the effectiveness, consistency, and completeness of system functions and internal controls [38,39].

Regarding the sustainability implications of DFSS, it has to be said that although lean manufacturing and sustainable development appear to be quite distant areas, in fact, they are closely related and interconnected [40]. The link between these two themes is represented by the fact that lean production is targeted at decreasing waste amounts by detecting and eliminating those activities with high inefficiencies and low added value to the product, thus achieving substantial benefits not only in operational terms but also under environmental, economic, and social perspectives [41]. In this context, the conception of an integrated and holistic DFSS-Sustainability approach could certainly support manufacturing activities and business plans for addressing the high costs related to the availability of primary resources and economic resources [42], and it would strengthen sensitivity towards social and environmental issues [43], thus acquiring a decisive and lasting competitive advantage [44]. That said, all redesign activities that derive from engineering bottlenecks identified by DFSS methodologies need to be evaluated from a sustainability point of view, for example, the re-design activity covering a wide range of interventions each one characterised by specific and sectoral LC effects [45]

This paper provides a practical application of DFSS to a real-life case study, focusing primarily on the identify phase of the IDOV process. The chosen case study is a specific mountain bike model (RG0-TT model) produced by the RGBike company [46], a middle-class product platform characterised by a durable and cost-effective aluminium frame. The adopted DFSS model is based on the implementation of methods presented in the introduction section according to the framework depicted in Figure A1 in Appendix A. The work includes the collection of different VOC types (reactive, proactive, literature and industry journal analysis), which are then screened and prioritised through the application of a series of tools (affinity diagram, market segmentation, analytical hierarchy process pairwise comparison). The study continues with the translation of VOCs into CTQs and the identification of relationships between VOCs and CTQs and between CTQs and CTQs, by means of the interaction matrix (roof matrix). Then, competitive benchmarking and definition of target objectives (VOC benchmarking from the customers' perspective and CTQ benchmarking) are carried out. Finally, the work provides prioritisation of CTQs and selection of critical characteristics for target achievement, based on which improvement strategies are identified and made available to the production company to apply product redesign.

2. Materials and Methods

The development of the DFSS case study is described below according to the following set of activities:

-

Drafting a simplified project charter;

-

Collection, analysis, and prioritisation of the VOC;

-

Compilation of QFD House of Quality (HoQ) to prioritise CTQ factors based on both previously calculated VOC weights and VOC-CTQ correlations;

-

Competitive benchmarking to compare the current state with hypothetical competitors.

Inventory data come from the literature, industry magazines, and market surveys.

2.1. Project Charter

The formal initiation of a project is formalised by the publication of a project charter, representing the formal act by which stakeholders are informed of all characteristics, tasks, responsibilities, and roles involved in the project itself. Such a document includes general project information, product purpose, company strategy, target markets, and available budget. For this case study, the business case and the related project charter are based on a market analysis published by Price Waterhouse Coopers (PWC) [47], concerning the global bicycle market. The main elements of the project scope are reported below.

-

Vision statement. Conquer any terrain with RG0-TT mountain bike model, the most versatile daily bicycle on the market, designed for those who do not want to compromise fun for total safety;

-

Business case. Market research shows that products offered by major competitors focus on (1) high-tech vehicles, with a high price range to satisfy highly engaged customers (5% of the market); and (2) low-tech vehicles, with a low price to satisfy predominantly travel-oriented customers (30% of the market). RGBikes aims to achieve market leadership by launching a new middle-class product, conceptually new and attractive, with the primary goal of ensuring the versatility of use. It is targeted at simultaneously satisfying three main categories of cyclists: travel, leisure, and engagement (representing together about 95% of the market). The new product must be highly customisable for both everyday use and more athletic use on any terrain. The technical basis of the bike allows the choice of two different types of products, mountain bike and hybrid/cross, which are the preferred segments by customers (about 50% of the total sold);

-

Objective and constraints. The main targets and requirements of the product are (1) a new product platform for hybrid/cross and mountain bikes; (2) the use of high-grade 7000 series aluminium frames; (3) a high-end braking system for improved safety; (4) high customisation possibilities; (5) the base equipment does not include a rear shock absorber, which remains a customisable option; and (6) a EUR 900 price cap.

2.2. VOC Collection

A specific plan dedicated to collecting the VOC is developed, which is divided into two parts. The first one is planning for reactive VOC collection, focused on warranty analysis and customer/dealer complaints. The second part is planning for proactive VOC collection, targeting both internal needs (needs of the production company and those of all stakeholders listed in the project charter) and external needs. The combined results lead to the collection of 42 VOCs, which are subsequently analysed, prioritised, and translated into CTQ characteristics.

Reactive VOC Collection. The collection of Reactive VOC is carried out by means of two main activities. The first one is the estimation of product reliability from the customer perspective, whose main purpose is calculating an index able to quickly establish the reliability level of a bicycle, which is then compared with project objectives. The second step is analysing and evaluating market issues and complaints, which are essential for continuous improvement in both existing and future products. The RGBike company has tested a sample of 15 high-end mountain bike vehicles with technical features comparable to the one under consideration in this paper. Each vehicle has been ridden for a total of 10,000 km, which is sufficient for the identification of infant mortality and the estimation of reliability during operational life (which, by design, is 10,000 km per year under a time span of 10 years). Figure 1 reports the cumulative failure rate for the anomalies, in terms of δ, which is the ratio between the number of cumulative failures and the number of vehicles under test [-], as a function of the mileage [km].

The cumulative failure rate of anomalies shows a rapid increase in the number of breakdowns occurring for low mileage values. After 1000 km mileage, the cumulative number of failures shows a linear trend that can be considered more representative of the entire useful life-time. On the other hand, the mileage ranges of 3400–4100 km, 6000–7000 km, and 8400–9000 km exhibit anomalous growths in the number of failures. A detailed analysis of the phenomena reveals that eight out of the fifteen vehicles are affected by anomalous wear of the front brake disc, with evidence of linear crack indications after 3400 km. Nine out of fifteen bikes provide an anomalous increase in the friction of the front fork (resistance to fork sinking) and an overall deterioration in vehicle handling at the 6000 km mileage. Six out of fifteen vehicles are affected by anomalous wear of the front brake disc, with evidence of linear crack indications after 8400 km of driving. The analysis reveals the following information that can be classified as reactive VOCs, particularly in regard to customer-specific requirements:

-

Friction of the fork and vehicle handling must remain consistent as the distance travelled increases;

-

The reliability of the brakes, under normal wear conditions, must not be compromised.

In the light of failure rate analysis, an investigation and evaluation of problems/complaints encountered in the market is carried out. The data collection of the highlighted problems is shown in

Table 1. Analysis of problems encountered in the market: comparison of absolute data and normalised data according to sales.

	Pure Data		Data Normalised by Sales
Functional Group	% Anomalies Europe	% Anomalies North America	% Anomalies Europe	% Anomalies North America
Brake System	34	66	25	75
Handlebar and Controls	47	53	37	63
Transmission System	39	61	30	70
Saddle System	65	35	55	45
Suspension System	34	66	25	75
Wheels System	33	67	25	75
Supporting Structure	0	100	0	100
% Bikes Sold in Europe	60
% Bikes Sold in North America	40

. The analysis reveals that the brake and handlebar assemblies are the ones with the highest priority for intervention. It also shows that, with the exception of the saddle assembly, the majority of anomalies occur on vehicles sold in North America, for which it is reasonable to hypothesise a more intensive or heavy use of the bikes.

The investigation is deepened by planning a proactive VOC collection with customers and dealers to find out the main usage patterns of the bicycles. Information is primarily collected through analysis of components that have shown anomalies or failures in the market, as well as customer care reports related to customer and/or dealer complaints. Regarding faults, a large proportion of the components have been subsequently analysed by the quality assurance department of the RGBike company. Complaints have been collected and sorted by the customer care department of the RGBike company. The results obtained are summarised in

Table A1. Breakdown of main faults and complaints by functional group.

	Failure	Complaint
Brake Group
Abnormal wear on the front brake disc with the presence of cracks	x
Noticeable pulsations on the lever during braking with the front disc		x
Noise during braking when activating the front brake		x
Oil leakage from the front brake pump, due to faulty gasket sealing	x
Reported low efficiency of the front brake during repeated braking		x
Reported low efficiency of the rear brake during repeated braking		x
Handlebar and Controls Group
Wear on the handlebar grips with multiple cuts	x
Reported discomfort with the grips and poor ergonomics of the handlebar during long rides (<50 km)		x
Fragile breakage of the lever of the rear gear shifter	x
Excessive force required on the lever to shift gears under load		x
Transmission System Group
Fragile breakage of the front derailleur	x
Noticeable oxidation on the cassette sprockets	x
Reported low grip on the pedal surface		x
Saddle Group
Wear on the microfiber contact areas of the saddle cover	x
Reported discomfort of the saddle after long rides (<50 km)		x
Limited adjustability of the saddle reported		x
Reported difficulty in adjusting the saddle		x
Suspension Group
Permanent deformation of the stanchions with consequent reduction in smoothness	x
Reported discomfort of the front suspension on rough terrain		x
Reported limited adjustability of the front suspension		x

of Appendix B.

The results stress that there is a clear correlation between the failure cases on the market, (i.e., wear of brake discs and deformation of front suspension stems and front suspension rods) and the results of the statistical analysis found during the functional testing of vehicles. The conversion of the above results into VOC is reported in

Table A2. Table A1 results converted in VOC.

VOC Brake Group
(1) High reliability of the brakes
(2) Braking should be smooth without pulsations at the lever
(3) Absence of noise during driving
(4) No oil leakage during braking
(5) High efficiency of the front brake
(6) High efficiency of the rear brake
VOC Handlebar and Controls Group
(7) The grips should not wear out
(8) The grips and handlebars should be comfortable (in all usage conditions)
(9) The gear levers should function correctly
(10) Ability to shift gears under load
VOC Transmission System Group
(11) The front derailleur should function correctly
(12) The sprockets should not rust
(13) The pedals should have sufficient grip
VOC Saddle Group
(14) The saddle should not wear out
(15) The saddle should be comfortable (in all usage conditions)
(16) The saddle should be sufficiently adjustable
(17) The saddle should be easy to adjust
VOC Suspension Group
(18) Low friction of the forks
(19) The suspension should be comfortable
(20) The suspension should allow for a wide range of adjustments

of Appendix B.

Proactive VOC collection. Proactive VOC collection is performed to understand the product through information coming from the customers, aiming at discovering latent needs, which are then categorised using the Kano Model [48,49,50]. To this end, field tests and events with representative customers have been carried out; these events are essential for direct interaction between the company and the consumers themselves, enhancing the brand prestige and attracting attention from potential customers. Various types of bicycles have been tested on three different circuits (asphalt, off-road, and cross-country), to identify the needs of the diverse use modes. This kind of activity has been performed by testing also some bikes of competitors, suitably camouflaged so as not to influence judgements, which have been made available to complete the survey. Targeted interviews are also conducted during the above events in order to:

-

Understand the requirements and performance levels that the product must have;

-

Assess customer priorities in terms of essential needs;

-

Know the must-have needs and be able to derive any “more is better features”;

-

Comprehend customers’ perceptions of competitors.

The interviews have been useful in preparing the end-of-activity survey, which has been completed online by the customers. The literature studies, scientific articles, and industry magazines are also taken into account; as a confirmation, a study conducted by Ayachi, Dorey and Guastavino on the perceived dynamic comfort of cyclists [51] is considered for insights into the VOC prioritisation phase.

Table A3. VOC Collection.

VOC
Category	Code
Reliability	1.1	The grips should not wear out
	1.2	The gear levers should function correctly
	1.3	The front derailleur should function correctly
	1.4	The sprockets should not rust
	1.5	The saddle should not wear out
Comfort	2.1	Smooth braking without pulsations at the lever
	2.2	No noise during braking
	2.3	High comfort of grips
	2.4	High comfort of handlebars
	2.5	Controls should be ergonomic
	2.6	The saddle should be comfortable (in all conditions)
	2.7	The suspension should be comfortable
	2.8	The riding position should be comfortable
Cost	3.1	The cost of the bicycle should not be excessive
	3.2	Maintenance costs should not be excessive
Design	4.1	The bicycle should have an appealing design
	4.2	The colour of the bicycle should be customizable
Efficiency	5.1	The grips should have good traction
	5.2	Low effort at the lever during braking
	5.3	Ability to shift gears under load
	5.4	The saddle should have good grips
	5.5	Low fork friction
	5.6	Provide good traction
	5.7	The bicycle should be responsive
	5.8	The bicycle should be lightweight
	5.9	The bicycle should have low rolling resistance
Services	6.1	Prompt after-sales service for issue resolution
	6.2	Long warranty period
	6.3	Spare parts should always be available
	6.4	Availability of accessories
Safety	7.1	High brake reliability
	7.2	High efficiency of front brake
	7.3	High efficiency of rear brake
	7.4	The pedals should have sufficient grip
	7.5	Good weight distribution
	7.6	Wheels should be impact-resistant
Versatility	8.1	The bicycle should adapt to body dimensions
	8.2	Saddle height should be easily adjustable while riding
	8.3	Saddle height should be sufficiently adjustable
	8.4	Suspension should be sufficiently adjustable
	8.5	Wheels should be easily removable
	8.6	The bicycle should be customizable

in Appendix B summarises the VOC collection results.

2.3. Prioritisation of VOC and Translation into CTQ

To construct the HoQ of QFD, the VOCs must be first prioritised and subsequently translated into CTQs. The requirements are prioritised by means of the Analytical Hierarchy Process (AHP) pairwise comparison, an advanced Multiple Criteria Decision Making (MCDM) method [52]. The AHP technique involves the following steps: (1) decomposition of requirements into a hierarchy based on affinities; (2) creation of the requirements correlation matrix; (3) comparison of requirements through pairwise comparison only among those in the same category; (4) calculation of the importance value within the category and between categories; (5) evaluation of decision quality through the consistency parameter; and (6) iteration of the process until an acceptable The Consistency Ratio (CR) value is obtained (CR < 0.1). Appendix C shows the method adopted for calculation, as well as the correlation matrices of the individual categories. The results of prioritising the main categories of VOC are summarised in

Table 2. Correlation Matrix of the main categories.

Correlation Matrix Categories
	Reliability	Comfort	Cost	Design	Efficiency	Services	Safety	Versatility
Reliability	1.00	2.00	2.00	5.00	2.00	2.00	1.00	3.00
Comfort	0.50	1.00	0.33	4.00	1.00	0.33	0.50	1.00
Cost	0.50	3.00	1.00	4.00	1.00	2.00	0.33	3.00
Design	0.20	0.25	0.25	1.00	0.33	0.25	0.20	0.20
Efficiency	0.50	1.00	1.00	3.00	1.00	1.00	0.33	1.00
Services	0.50	3.00	0.50	4.00	1.00	1.00	0.33	3.00
Safety	1.00	2.00	3.00	5.00	3.00	3.00	1.00	5.00
Versatility	0.33	1.00	0.33	5.00	1.00	0.33	0.20	1.00
Sum	4.53	13.25	8.42	31.00	10.33	9.92	3.90	17.20

and

Table 3. Analysis of problems encountered in the market: comparison of pure data and normalized according to sales.

Normalised Correlation Matrix Cat. Comfort
	Rel.	Com.	Cost	Des.	Eff.	Ser.	Saf.	Ver.	Importance	C_Measure
Rel.	0.22	0.15	0.24	0.16	0.19	0.20	0.26	0.17	0.20	8.58
Com.	0.11	0.08	0.04	0.13	0.10	0.03	0.13	0.06	0.08	8.20
Cost	0.11	0.23	0.12	0.13	0.10	0.20	0.09	0.17	0.14	8.83
Des.	0.04	0.02	0.03	0.03	0.03	0.03	0.05	0.01	0.03	8.30
Eff.	0.11	0.08	0.12	0.10	0.10	0.10	0.09	0.06	0.09	8.54
Ser.	0.11	0.23	0.06	0.13	0.10	0.10	0.09	0.17	0.12	8.69
Saf.	0.22	0.15	0.36	0.16	0.29	0.30	0.26	0.29	0.25	8.75
Ver.	0.07	0.08	0.04	0.16	0.10	0.03	0.05	0.06	0.07	8.27
CI	0.07
CR	0.05

.

As shown, the most important categories are “safety” and “reliability”, closely followed by “cost” and “services”. The importance of individual VOCs is then normalised to the importance of the belonging category. Based on the results of [51], the company decided to assign an additional importance value (+1.5%) to VOCs related to the handlebar, saddle, and comfort of the riding position. The frame material (aluminium) is a project constraint for cost containment. For each VOC, one or more CTQs are identified to provide an appropriate description of VOCs from a technical point of view by means of measurable parameters, as shown in

Table A20. Conversion of VOC in CTQ.

Category	VOC Code	VOC	CTQ
Reliability	1.1	The grips should not wear out	MTTF Rubber Hardness of handlebar grips
	1.2	The gear levers should function correctly	MTTF number of gears
	1.3	The front derailleur should function correctly	MTTF number of gears
	1.4	The sprockets should not rust	Salt spray resistance
	1.5	The saddle should not wear out	MTTF Abrasion resistance
Comfort	2.1	Smooth braking without pulsations at the lever	DTV of the front brake disc
	2.2	No noise during braking	Noise level
	2.3	High comfort of grips	Grip comfort
	2.4	High comfort of handlebars	Handlebar comfort
	2.5	Controls should be ergonomic	Control comfort
	2.6	The saddle should be comfortable (in all conditions)	Saddle comfort Foam thickness Foam density Δ foam thickness under load
	2.7	The suspension should be comfortable	Suspension comfort Maximum fork travel Spring stiffness Maximum adjustment of static Support And Gear (SAG)
	2.8	The riding position should be comfortable	Riding position comfort; Seat-handlebar distance; Seatpost-bottom bracket distance;
Cost	3.1	The cost of the bicycle should not be excessive	Cost of Handlebar Group Cost of Controls Group Cost of Suspension Grouo Cost of Braking System Group Cost of Supporting Structure Group Cost of Saddle Group Cost of Transmission Group Cost of Wheels Group
	3.2	Maintenance costs should not be excessive	Maintenance cost Spare parts cost
Design	4.1	The bicycle should have an appealing design	Design evaluation
	4.2	The colour of the bicycle should be customizable	number of available colours
Efficiency	5.1	The grips should have good traction	Grip evaluation of handlebar grips Static and dynamic friction coefficient of handlebar grips
	5.2	Low effort at the lever during braking	Maximum lever effort
	5.3	Ability to shift gears under load	Maximum shift load number of gears
	5.4	The saddle should have good grips	Saddle grip evaluation Static and dynamic friction coefficient of the saddle cover
	5.5	Low fork friction	Maximum axial friction load Maximum friction load with caster
	5.6	Provide good traction	Total weight of the bicycle Tire grip coefficient Maximum fork travel Static weight of front wheel Static weight of rear wheel number of gears
	5.7	The bicycle should be responsive	Total weight of the bicycle Yaw stiffness of frame Torsional stiffness of frame Bending stiffness of frame Weight of front wheel Weight of rear wheel
	5.8	The bicycle should be lightweight	Weight of Handlebar Group Weight of Controls Group Weight of Suspension Group Weight of Braking System Group Weight of Supporting Structure Group Weight of Saddle Group Weight of Transmission Group Weight of Wheels Group
	5.9	The bicycle should have low rolling resistance	Aerodynamic drag coefficient Power dissipated by rolling resistance Power dissipated by transmission friction Total weight of the bicycle
Services	6.1	Prompt after-sales service for issue resolution	MTTR
	6.2	Long warranty period	number of warranty years
	6.3	Spare parts should always be available	Spare parts availability time
	6.4	Availability of accessories	number of available accessories
Safety	7.1	High brake reliability	MTBF/MKBF of brakes
	7.2	High efficiency of front brake	Maximum deceleration Total weight of the bicycle Maximum fork travel Front brake disc size Static weight of front wheel Static weight of rear wheel
	7.3	High efficiency of rear brake	Maximum deceleration Total weight of the bicycle Static weight of front wheel Static weight of rear wheel Rear brake disc size
	7.4	The pedals should have sufficient grip	Pedal surface area
	7.5	Good weight distribution	Wheelbase Seat-handlebar distance Static weight of front wheel Static weight of rear wheel
	7.6	Wheels should be impact-resistant	Rim safety coefficient; Front fork travel; Front rim diameter; Rear rim diameter;
Versatility	8.1	The bicycle should adapt to body dimensions	number of sizes
	8.2	Saddle height should be easily adjustable while riding	Seat height adjustment time
	8.3	Saddle height should be sufficiently adjustable	Maximum seatpost length
	8.4	Suspension should be sufficiently adjustable	Maximum fork travel; Maximum adjustment of static SAG
	8.5	Wheels should be easily removable	Wheel removal time
	8.6	The bicycle should be customizable	number of available customizations

in Appendix D.

2.4. Quality Function Deployment 1

Following the prioritisation of VOC and the translation into CTQ, the subsequent step of the work is completing the HoQ of QFD1. For this task, the Qualica 19 software is used [53], which supports a wide range of tools used in the DFSS approach. The methodology involves evaluating the correlations between the previously prioritised VOCs (42 elements) and the CTQ characteristics, appropriately selected (79 elements). A total of 3318 comparisons are made, and for each one, an attempt is made to answer the question “Which effect on meeting Need Y is caused by an improvement in CTQ X?”. As provided in Figure 2, a logical sequence is then assigned to CTQs based on their relative importance, in descending order: items with the highest values are those that must be satisfied first in the design phase. Figure 2 also highlights correlations between the top six VOCs and the top ten CTQs, using different colours. The power and effectiveness of QFD are demonstrated by the fact that there is no one-to-one correspondence between CTQ and VOC based on the importance level. The tool allows the identification of technical characteristics that contribute to maximising customer satisfaction. Thanks to the roof of the quality house, it is possible to analyse correlations between various CTQs, to assess the existence of positive or negative interactions. For each cell of the roof matrix, an answer is found to the question “What effect does a possible improvement in CTQ A have on CTQ B? Positive, negative, or none? And if there is an effect, how strong is it?”. The total number of comparisons made is 6162. The results of the analysis are reported in Figure 3.

The quadrants in Figure 3 divide CTQs into four categories:

-

Passive CTQs (grey quadrant), whose active and/or passive (positive or negative) relationships are weak;

-

Reactive CTQs (yellow quadrant), that have the majority of passive relationships, whether positive or negative;

-

Active CTQs (violet quadrant), which provide the majority of active relationships, whether positive or negative;

-

Critical CTQs (red quadrant) that reveal a large number of both active and passive relationships, whether positive or negative.

A further investigation into the type of relationship, whether positive or negative, highlights that:

-

The priority for improving CTQs 2.8, 8.1, and 8.2 needs to be increased;

-

A detailed analysis of CTQs 3.10, 6.2, 7.3, and 8.5 is necessary: contradictions can be resolved using the methodologies outlined in the introduction section.

2.5. Competitive Benchmarking

In the process of developing a new product, a comparison with the main market competitors is required, with the target to not only create a clear and concise picture of strengths and weaknesses from the customers' perspective but also to identify critical improvement areas. Competitive benchmarking is performed, including VOC benchmarking, where product performance is evaluated in terms of the customers’ satisfaction, which provides an assessment of the product based on CTQ metrics. Evaluations of two of the main market competitors have been collected through direct measurements, field tests and analysis of brochures/technical data sheets. Products belonging to the same category are compared, with price ranges consistent with the company strategy. The complete set of results from VOC benchmarking and CTQ competitive benchmarking is reported in Figure A2 and Figure A3 in Appendix E, which include the following:

-

Target values for each CTQ, these latter are based on results of VOC benchmarking and measurements;

-

Engineering difficulty levels, which are assigned on a (1–10) scale, to highlight CTQ engineering bottlenecks that are important but difficult to implement (Importance > 5; Difficulty > 5).

3. Results and Discussion

This section provides the main results of this study, along with a critical discussion of them.

First of all, products belonging to the same category are compared, with price ranges consistent with the company strategy. The CTQ baseline (reported in Figure A3 in Appendix E) provides a score of 5.1, which results are in line with Competitor 1 (4.9), but it is distant from both the evaluation of Competitor 2 and the target (respectively, 5.5 and 6.8).

Table 4. CTQs to improve ranked from the highest importance.

CTQ	Unit	Baseline	Target Value	Difficulty	Importance AHP (%)	Importance (0–10)
7.1 MTBF/MKBF Brake Group	km	3500	8000	5	3.66	10
8.2 Seat Height Adjustment Time	s	60	3	7	3.52	9.6
8.5 Number of Available Customizations	# Costom.	4	8	3	2.67	7.3
7.3 Front Brake Disc Size	mm	140	180	1	2.65	7.3
7.2 Maximum Deceleration	g	0.32	0.4	3	2.53	6.9
1.3 MTTF of Front Derailleur	h	160	300	3	2.47	6.8
6.2 Number of Warranty Years	# years	2	3	5	2.23	6.1
1.1 MTTF of Handlebar Grips	h	140	300	5	2.02	5.5
7.4 Rear Brake Disc Size	mm	140	160	1	1.89	5.2
1.6 Number of Gears	# gears	8	10	1	1.61	4.4
1.7 Salt Spray Resistance	h	24	48	3	1.41	3.9
1.4 MTTF of Saddle	h	145	300	3	1.40	3.8
3.1 Cost of Handlebar Group	EUR	32	22	3	1.35	3.7
3.2 Cost of Control Group	EUR	46	40	3	1.35	3.7
3.3 Cost of Suspension Group	EUR	73	60	2	1.35	4.5
5.8 Maximum Friction Load with Caster	N	109	70	4	1.24	0.6
1.8 Abrasion Resistance	# rubs	67	100	4	1.15	3.1
4.2 Number of Available Colours	# colours	-	7	1	0.69	1.9
2.9 Foam Thickness	mm	15	25	1	0.58	4.8
5.7 Maximum Axial Friction Load	N	88	50	4	0.52	0.6
2.10 Foam Density	g/cm3	-	1.45	1	0.44	1.4
5.23 Aerodynamic Drag Coefficient	m2	0.38	0.4	3	0.22	3.7

reports the CTQs that need improvement to reach the target value. CTQs are ranked from the highest importance value to the lowest value. Six CTQs are from Group 1 (Reliability), four CTQs from Group 7 (Safety), three CTQs from Group 3 (Cost), and three CTQs from Group 5 (Efficiency). The first two CTQs, 7.1 and 8.2, are critical as they are characterised by a high importance value (3.66% and 3.52%, or 10 and 9.6 on a scale of 0–10), and at the same time by a notable difficulty value (5 and 7).

Figure 4 reports the CTQs in relation to engineering difficulty level and importance level. Based on these results, the CTQs located in the critical quadrant (first quadrant) are identified as engineering bottlenecks, on which efforts of a possible redesign activity should be concentrated to improve customers’ appreciation for the product. Such critical CTQs are the following:

-

C8.2: saddle height adjustment time;

-

C7.1: brake System Mean Time Between Failures (MTBFs)/Mean Kilometre Between Failure (MKBF);

-

C6.2: number of years of warranty;

-

C1.1: handlebar grips Mean Time To Failures (MTTFs).

Analysis of CTQ engineering bottlenecks and HoQ roofs suggests increasing the priority level of CTQ 8.2 implementation, along with further investigation of the relationships of CTQ 6.2. While the remaining CTQs do not show critical relationships with other vehicle features, their improvement results are challenging. The use of problem-solving techniques presented in the previous sections can help to define and break down problems, as well as to resolve contradictions. Some CTQs are considered “Quick Win”; these CTQs (reported in Figure 4) can be improved relatively easily and they are considered of high importance. For example, C7.3 “Front Brake Disc Size” and C7.4 “Rear Brake Disc Size” must reach the target value of 180 mm and 160 mm, respectively, while the diameter for both baseline components measures 140 mm. At the same time, these CTQs have an importance value of 2.65% and 1.89%, respectively (7.3 and 5.2 on a scale of 0–10). Another example of “Quick Win” CTQ is C1.3 “MTTF of Front Derailleur”, which should reach the target value of 300 hr, while currently it amounts to 160 hr. This CTQ has an importance value of 2.65% (7.3 on a scale of 0–10), and the engineers evaluated his improvement as easy to implement, rating the difficulty at 1.

In regards to the environmental implications caused by the application of DFSS, the resolution of engineering bottlenecks (as well as critical CTQs) does not necessarily mean redesign interventions, which involve a modification of the product's environmental profile. Considering engineering bottlenecks identified, an example of this is represented by CTQ C8.2 “Saddle height adjustment time”, whose improvement is not strictly related to a variation in the LC inventory for both the saddle and the head tube components. On the other hand, redesign interventions applied on the basis of CTQs referred to MTBS and MTTF can surely involve a change in the amount of

-

Material and energy resources required to produce or maintain a specific component (or a group of components);

-

Emissions and wastes produced at End-of-Life (EoL) of different bike parts.

Indeed, increasing MTBS and MTTF can be actually achieved by a change in the components’ material or in the components’ design, with the latter possibly involving a modification in the manufacturing processes. That said, a variation in material, design, or production technology always causes a modification in the number of elementary flows (that is primary resource flows and substances emitted into the environment), which determines the overall environmental effects of a specific LC stage (or a specific section of an LC stage, such as a production process or an EoL treatment) of the final product. Assuming, for example, that design modification consists of a change in material for one or more components (for example, an aluminium head tube that is replaced by a composite head tube because of its lower need over time for maintenance interventions), it has to be considered that such a material variation involves multiple and interconnected effects on the sustainability profile of the entire bicycle. First of all, a switch from aluminium to Carbon Fiber Reinforced Plastic (CFRP) would result in the vast majority of cases in an increased impact on the production stage. This is due on the one hand to the notably larger damages (i.e., Global Warming Potential, GWP) in the manufacturing of semi-finished products of carbon fibres when compared to aluminium, mainly due to the strongly higher energy intensity of extraction and refining processes of minerals and base materials. On the other hand, also the production of semi-finished products into the final component is often definitely environmentally detrimental when considering innovative materials (especially composites), primarily due to the vastness of processes and energy/material consumables necessary for their manufacturing. Looking at the other LC phases, CFRPs appear to be also not eco-friendly when considering the disposal at EoL, due to notable technical difficulties in the separation of different constituent materials. At the same time, aluminium can be easily separated and also conveniently recycled; as a confirmation, the recycling of aluminium is characterised by a high substitution factor (net quota of primary raw materials whose extraction is avoided thanks to recycling), which is approximately 40% [54], even larger than the one for steel (30%) [55,56]. On the other hand, lightweighting (which can be achieved also by improving CTQs that do not fall within engineering bottlenecks—such as CTQ C5.13 “Mass of bike frame”) can offer significant advantages in the use stage, but only when the vehicle consumes energy for its propulsion. As a consequence, in the present case study, weight reduction can be considered an effective way to improve the environmental profile of the product only when improvement strategies identified downstream of DFSS provide for vehicle electrification (or in the case that this entire study is conducted on an e-bike model). In such a case, lower mass means fewer effects associated with the production of a reduced amount of energy required for traction, and such a benefit increases the bicycle LC distance.

In light of previous considerations, it has to be stressed that possible design modifications downstream of the DFSS analysis do not necessarily involve a reduction in the environmental profile of the mountain bike, as a complex assessment is required to properly evaluate their consequences, that is, advantages in specific LC phases/processes and disadvantages in certain other. Additionally, it is important to consider that the sustainability evaluation of product LC should not be limited to the environmental sphere, but it should extend to economic and social aspects, by means of Life Cycle Costing (LCC) and Social Life Cycle Assessment (S-LCA) methodologies, respectively. Obviously, the combined implementation of all these different types of analyses requires a very strong effort in data collection (both in terms of time and personnel), due to the substantial amount of necessary information.

The main limitation of this study is the fact that an actual product redesign on the basis of engineering bottlenecks identified by the DFSS analysis is missing, which represents the most logical continuation of the work in future developments.

4. Conclusions

This paper presents the implementation of the DFSS method in a real-life case study, a bicycle. The main target of the work is identifying applicable improvement solutions to the baseline version of a specific mountain bike model produced by the RGBike company [41]. The collection and prioritisation of VOCs, their conversion to CTQs, the definition of relationships through QFD1, and the analysis of VOC benchmarking and CTQ benchmarking are described in detail. The paper also presents the prioritisation of CTQs and critical target selection, identifying the main product engineering bottlenecks that mainly need improvement by application of redesign interventions.

The case study results show that defining a universal product development process capable of equally addressing all design problems is not possible, while it is advisable to establish a structured guideline that allows identifying the most appropriate problem-solving methodology for the specific situation: DFSS methodologies and analysis tools must be carefully selected and adapted based on context, product type, and experience of practitioners. Results also stress that when the complexity level of the considered product increases (in terms of, for example, the number of components, materials used, and manufacturing processes, etc.), a synergistic use of available DFSS methods is fundamental for achieving the maximum efficiency. In this regard, the authors conclude that a combined DFSS approach is the best strategy for using seemingly distant methodologies to enhance the overall product performance level, while at the same time improving business processes and achieving production excellence.

Nowadays, there is a growing recognition of DFSS’s potential to drive sustainable development (in contrast to previous years, where the focus of DFSS was predominantly on operational efficiency), and the results are also critically discussed from a sustainability point of view. Based on the analysis of engineering bottlenecks identified for the mountain bike case study, possible redesign interventions can surely affect the environmental profile of the product itself, but such a variation does not necessarily mean a decrease in the impacts. This is mainly due to the fact that design changes (in terms of materials, design solutions, manufacturing processes, and technologies) involve a series of contrasting effects on different product LC stages, the proper evaluation of which requires a structured assessment able to capture all the above mentioned conflicting consequences. The authors finally highlight the implications of possible design modifications downstream of the DFSS analysis should be assessed from a holistic sustainability perspective by combining DFSS with LC analysis methodologies (such as Life Cycle Assessment, Life Cycle Costing, and Social Life Cycle Assessment), with these latter appearing as a promising area for future research.