Practical Approach for Smart and Circular Cities: Chatbots Used in Waste Recycling

Abstract

Sustainable development, smart waste management, and circular economy principles are paramount to the significant worldwide trend of smart city-related research and projects. The basic hypothesis of our research is that artificial intelligence (AI)-based IT applications have an increasingly important role in the field of smart cities in terms of issues related to waste management. In our present article, we set out to analyze the characteristics of chatbot applications dedicated to waste recycling in the case of smart cities and propose some innovative ideas to improve the efficiency of such applications. Based on the consultation and analysis of a whole series of chatbot-type applications used to facilitate the recycling activity, we systematically analyze and evaluate five illustrative examples of chatbots employed in the context of material recycling. We provide performance comparisons in a table based on specific relevant criteria. Furthermore, the detailed analysis of these chatbots has led to the idea of improving the performance of this type of application. In this sense, we propose a series of innovative concepts that can be successfully implemented in future chatbots dedicated to the field of the circular economy. Here, we detail the innovative ideas that can promote the circular economy and capitalize on the potential of chatbots in the waste recycling activity. We also identify some possible limitations of these new ideas that we propose to be implemented. As for future research directions, our goal is to develop a chatbot dedicated to improving waste recycling practices within the framework of a smart city. Such innovation holds promise in improving sustainability efforts and fostering environmental stewardship within urban environments.

1. Introduction

In 2050, approximately 70% of the world’s population is estimated to live in the urban environment [1], so the urban agglomerations formed will need an intelligent infrastructure that can optimally manage the needs of citizens and offer increasingly diverse and more efficient services [2].

In this sense, the concept of the circular economy (CE), in which resources must be used for as long as possible and saved, has attracted special attention in recent years to achieve sustainable (urban) development. In this context, the digital CE represents a new approach through which digital technologies can enable the development of a more circular and regenerative global economy. Within the framework of the CE, the generation of waste is systematically curtailed through meticulous product design and an industrial process characterized by the perpetual circulation of materials within a “closed-loop system”. This paradigm prioritizes sustainability by minimizing the linear progression from resource extraction to disposal, promoting a regenerative and resource-efficient approach to production and consumption [3].

Furthermore, in this context, the implementation of waste management and recycling is of paramount importance in the dynamic landscape of smart cities, where the intersection of urbanization and advanced technologies requires a sophisticated approach to environmental management. As urban centers face increasing volumes of waste, the need for sustainable waste practices becomes key to the ecological resilience of smart cities [4]. Moreover, the ubiquity of waste has reached a global scale, with an unprecedented demand for raw materials. Statistics indicate that approximately 80% of all materials and consumer goods are subject to disposal, contributing to an increasing challenge. Furthermore, within the intricate web of the food supply chain, an alarming excess of more than 30% of processed food is discarded, underscoring the urgent need for comprehensive waste management strategies [5].

The hypothesis of our research posits that AI-driven IT applications play a pivotal role in smart-city contexts, particularly in addressing waste management challenges. Therefore, the incorporation of AI technologies emerges as a strategic imperative to optimize waste processes, promote resource efficiency, and ultimately contribute to the overall goal of sustainable urban development [6].

Building on our previous research endeavors, which resulted in the publication of scientific articles on this topic [7,8], the present research study aims to examine five illustrative examples of chatbots employed in the field of waste management, with a specific focus on material recycling. In addition to our previous work, we posit certain directives and methodologies capable of enhancing the efficacy of software implementations of this nature.

In the concluding section of this article, we derive comprehensive insights, underscoring the pivotal role of chatbots in the circular economy. Additionally, we describe their practical implementation, culminating in a set of conclusions that highlight the importance of these intelligent systems in fostering sustainable practices and circular economy initiatives.

2. Literature Review

In recent years, the integration of AI technologies, particularly chatbots, has emerged as a promising solution to address waste management challenges in the context of the circular economy. In essence, a chatbot represents a sophisticated computer algorithm engineered to emulate natural conversation with human users, typically facilitated through text or voice interfaces. This review of the literature aims to provide a comprehensive analysis of five distinct recycling chatbots, highlighting their functionalities, strengths, and limitations based on the available literature and case studies.

In general, a comparative analysis reveals that while each recycling chatbot offers unique functionalities and strengths, they all collectively contribute to the promotion of sustainable waste management practices within the circular economy. By leveraging AI technologies, these chatbots facilitate user engagement, provide relevant information, and streamline recycling processes, thus fostering environmental awareness and participation at the individual and community levels. However, challenges such as data accuracy, customization, and user accessibility remain pertinent considerations for further development and implementation of recycling chatbots in real-world scenarios.

According to a study by EdTechReview of the recent development of the startup 1MillionBot, with a focus on securing a financial infusion of EUR 1.5 million, the company is valued at EUR 8.2 million. Founded in 2018 in Alicante by its founder Andrés Pedreño, the commercial entity develops specialized chatbots for various sectors and integrates them into various platforms, including Facebook Messenger and Google Assistant. In particular, this review highlights the company’s efforts to achieve precision and efficiency in the field of virtual interactions, with applicability in contexts as diverse as financial, academic, and administrative local institutions [9].

In contrast, the second article provides a general overview of 1MillionBot, highlighting its contribution to the development of virtual assistants in various public and private spheres. Mention is made of a previous round of funding, amounting to EUR 1.5 million, without detailing the exact use of these resources or the company’s development prospects. Furthermore, emphasis is placed on the presence of a department dedicated to research and development in the field of AI, along with monitoring technological progress in this area, but specific details about the product portfolio or strategic partnerships implemented by the company are missing from the academic presentation [10].

In one of the articles written by Gilroy Menezes, Vice President of Kilowott, a Goa-based tech startup, a detailed analysis of the evolution of chatbots and their potential to address environmental issues was provided. Menezes highlights the significant impact of chatbots on waste and pollution management, highlighting the case of Plastic Reporter, the first chatbot launched by WWF India in Goa. This chatbot allows citizens to report information on plastic waste through the WhatsApp platform, representing a key innovation in waste management [11].

Another article, published in the Times of India, details the launch of the chatbot, highlighting its importance in the context of an urgent need for effective waste management solutions in Goa. This chatbot, known as the Plastic Reporter, is an innovative intervention in the global fight against waste and pollution, empowering citizens to report and manage plastic waste information effectively. The article highlights the crucial role of chatbots in mobilizing citizens and promoting more responsible waste management practices, highlighting the importance of technological innovation in tackling environmental issues [12].

In one of the articles written by Sasha Dmytruk, published on Liga.net, the author reveals the transformation of No Waste Ukraine from a non-profit organization to a business with a social impact, addressing its evolution from a waste-sorting initiative to a sustainable business model. The article highlights the necessity of this transformation for the organization’s financial survival, emphasizing the importance of adapting to market conditions and socioeconomic challenges such as the pandemic and the invasion of Russia. It also outlines how No Waste Ukraine is diversifying its income through commercial services and new initiatives, such as opening its own shop and a sorting pod [13].

In another article published on Rubryka about the No Waste Ukraine chatbot, they focus on recognizing the efforts of companies in Ukraine that adopt responsible environmental practices, highlighting No Waste Ukraine’s initiative to award these companies with the No Waste Workplace distinction. The article highlights the organization’s active involvement in promoting a culture of waste sorting in the business environment and emphasizes the importance of corporate social responsibility in waste management. By presenting concrete examples of the companies who received the award, the article illustrates how No Waste Ukraine contributes to the formation of a sustainable waste management culture within the private sector [14].

3. Materials and Methods

In our research, we analyzed the first 100 research articles published in the years 2023–2024 by searching Google Scholar based on the keywords “waste management smart cities” (in total, the search returned a total of 4120 results). Therefore, we have observed different investigations dedicated to the analysis of recycling practices implemented at the municipal level in global cities.

When talking about recycling activity in a city, for example, in Orlando, Florida (USA), some notable transformative results have been achieved through the synergy of local policy and educational initiatives. In this direction, in 2019, the city implemented an ordinance mandating recycling practices in its 75,000 multifamily residential units. Collaborating closely with the municipality, this partnership secured two grants, the first in 2019 to facilitate the seamless implementation of the recycling ordinance, and the second in 2021 to extend the program’s reach to too low- to moderate-income housing. During the course of both initiatives, the program successfully expanded recycling access to almost 50,000 homes, resulting in the diversion of an additional 1200 tons of recyclables per year [15].

In another vein, we must emphasize the fact that recycling is important for reducing waste, because “waste is only waste when it cannot be reused, or its economic value is not economically feasible to become a resource” [16]. Furthermore, due to the recent progress in waste management, the use of first-hand materials can be reduced by 25% by 2030 [17].

Continuing the search for existing chatbot solutions in the world related to the topic of recycling in big cities, we found five products of this type, which we have described, analyzed, and compared in our article. Next, analyzing the chatbots presented, we have proposed a series of concepts that we consider innovative to improve AI applications (in other words, chatbots) designed to support the health of the environment. In general, these proposals refer to three main directions:

• Consumption and recycling tracking;
• Personalized advice;
• Education and awareness.

The following methods were used in the making of this article:

• Methods based on qualitative research;
• Methods that are based on quantitative research;
• Methods based on qualitative research and quantitative research (mixed methods).

Five chatbot models implemented in the circular economy in European countries or globally were described, analyzed, and compared. The smart solutions presented were chosen following a careful selection process, in which we particularly focused on the following:

• The originality of the chatbot;
• The field of activity in which the chatbot was developed and implemented;
• The benefits of using the application in the circular economy, but also in other sectors of activity;
• The number of users of the application;
• If the chatbot provides correct and relevant information;
• If the application is easy to use;
• Security and privacy of user data;
• Improvements that can be made to increase efficiency in the use of these chatbot models.

The qualitative research procedure was based on the use of the following methods:

• Case studies;
• Participant observations;
• Focus groups;
• A thematic study of documents.

Using case studies, we conducted a search, informed ourselves and identified chatbot models used in the field of the CE. The options implemented lead to an increased positive impact on their use, both in the chosen field and in the sectors of activity where the products are recycled and reused. We identified the benefits of these solutions for the companies/communities that implemented them, what technological functionalities these models developed, what security problems they faced, how the chatbot managed to achieve an effective dialogue with the user, and how to process and interpret images related to recycling questions. The method was used the most in the first part of our research, during which we carried out the following:

• We documented on the Internet countries that have understood the need to use technology in the field of the circular economy and are successfully implementing original solutions;
• We searched for, identified, watched, and analyzed videos showing different chatbot models implemented in this field;
• We read articles and specialized magazines published in the country or abroad about online platforms in the field of the circular economy which had integrated a chatbot;
• We identified the impact of recycling and the reuse of materials in the areas that chose to use the technology;
• The identified documents were carefully filtered, keeping only the relevant materials.

The discussions were based on different chatbot models implemented in the CE in European countries or around the world. The addressed issues revolved around subjects such as the following:

• Challenges in communities related to waste management;
• Innovative approaches that allow users to send photos/images of different types of waste to the chatbot;
• Help, advice, or recommendations given by the chatbot for the proper sorting of waste;
• Adapting intelligent solutions to the needs of the community (customizing the chatbot);
• Expanding the applicability of the technology also in the marine/oceanic space (the GOA Chatbot presented in this article puts particular emphasis on waste disposal located in marine ecosystems);
• Routing a user’s call to a human agent when the chatbot cannot provide the desired data based on user sentiment analysis (SCC Chatbot);
• The ability of a chatbot to handle multiple dialogs or calls simultaneously.

The focus group consisted of the following:

• The permanent and sustained collaboration between the authors who carried out this research;
• Collaboration with senior colleagues who work in the IT field and are interested in implementing information technology in the circular economy;
• Permanent appreciation, encouragement, positive feedback, and support from the members of the groups defined above, constantly referring to the goals and objectives which are pursued.

The thematic study of the documents consisted of the following:

• Identifying, reading, and analyzing specialized articles published both in the country and abroad about chatbot models developed and implemented in the CE;
• Studying other chatbot models implemented in online platforms and observing the tests conducted before the implementation of the model, the improvements made to the chatbot or the platform, the additional options added to the application, and a dialogue based on clear and correct information regarding the recycling activity;
• Watching videos about chatbot models implemented in various communities in the circular economy and their benefits.

The quantitative research procedure was based on the use of the following methods:

• Correlational studies;
• Structured observation;
• A quantitative study of the documents.

Through the method of correlational studies, information was collected that helped us identify, present, analyze, and compare chatbot models developed and implemented in the circular economy in different countries in Europe or the world. The data provided strong arguments about the need to use AI in the circular economy, to educate people and raise awareness of environmental issues, to manage waste, and to improve recycling behavior in the community. Evidence of the positive impact of AI in the circular economy in countries around the world is also supported and presented in articles produced individually or collaboratively by each member of the working group. Useful information is also collected from other studies and articles by experts in the field who have argued for the use of chatbots in the circular economy.

Structured observation is the method by which we collected information on chatbot models implemented in the circular economy. Data were gathered from the official websites dedicated to the chatbots. The chatbots chosen as examples of AI approaches to waste recycling challenges differ by offering customized solutions.

1MillionBot and LitterBot are distinguished by the integration of questionnaires that comprise questions that the chatbot asks the user to find out the material of which the product to be recycled is made. The questions also include information on a recycling number found on the product packaging. The purpose of these questions is to inform the user if they can recycle the waste.

SCC Chatbot stands out for its ability to forward complex queries to human agents.

GOA Chatbot differentiates itself with its innovative approach to image processing by allowing users to submit photos of plastic waste.

No Waste Ukraine has innovative solutions to help sort waste in regions with limited infrastructure by adapting solutions to the specific needs of the community.

Each chatbot uniquely collaborates with specific platforms such as call services, Telegram, or WhatsApp. Although they are different, the five chatbot models focus on providing accurate and relevant information about recycling practices. This technology raises community awareness and involves the adoption of sustainable practices within the circular economy.

The quantitative document study is the method by which we searched and gathered information about chatbot models developed and implemented in the circular economy. Data were collected from academic articles presented at international conferences or published in prestigious journals and were also inspired by courses presented in the research activity of the Doctoral School of Economic Informatics of the Bucharest University of Economic Studies. The research process carried out by all members of the working group also included the use of methods based on qualitative and quantitative research (mixed methods). The mixed-method strategies that we used in the investigation were sequential mixed-method procedures. These procedures were carried out in several stages. The design was based on an explanatory sequential design. The stages of the procedures are presented in Figure 1.

All references used in this research study were selected from prestigious journals, which contain recent studies and qualitative analyses with concrete information about the development and implementation of the chatbot in the field of the CE, as well as its relationship with a circular economic environment.

4. An Analysis of Five Representative Chatbots That Support the CE

Among the myriad of applications of AI lies the chatbot, meticulously crafted to engage and communicate with humans through automated interactions. Tailored to specific implementations, these chatbots serve as informative companions, providing users with valuable information on diverse topics while incorporating multifaceted functionalities to elevate the overall user experience. An illustrative instance is the integration of the ‘entity recognition’ pattern, adept at identifying and differentiating various objects. This functionality, when applied judiciously, contributes significantly to the circular economy’s objective of refining waste sorting and recycling processes, thus fortifying their efficacy and sustainability.

The prospective impact of chatbots in the future is of profound significance. Users, rather than relying on conventional websites or search engines for information retrieval, can now participate in human-like interactions facilitated by intelligent chatbots at various stages of inquiry. This potential extends to the sphere of the circular economy. To underscore this dimension and exemplify current market offerings, this study focuses on the analysis of five specific chatbots. Each of these chatbots exhibits common functionalities, coupled with domain-specific features tailored to their respective circular subdomains, whether it is about energy conservation, recycling, or other related aspects.

Without further introduction, the subsequent section delineates a comprehensive examination of five noteworthy chatbots specifically designed to facilitate citizens in large urban centers with the process of material recycling.

4.1. 1MillionBot

1MillionBot, founded in 2018, is an AI technology company known for developing virtual assistants (AVIs) and custom AI solutions. With more than 120 successful projects nationally and internationally, the company has established itself as a leader in this field in the public and private spheres. Through organic and inorganic growth, 1MillionBot has formed the 1MillionBot Group, which operates in various domains, including Software as a Service (SaaS), universities, local councils, governments, companies, and professionals [18].

One of 1MillionBot’s strengths is its AI research and development department, which has successfully implemented projects funded by prestigious institutions such as the CDTI (Center for Industrial Technology Development). This department closely follows developments in AI and natural language processing (NLP), adapting its solutions to integrate cutting-edge technologies and add value to ongoing projects [18].

It offers various solutions, products, and services in AI and virtual assistants, from custom development to integration with existing systems and regulatory compliance. The company has developed proprietary technologies focused on LLM models, the Millie platform to manage intelligent assistants, and unique SaaS products to democratize AI. This chatbot can improve and streamline waste collection through AI technology by providing users with personalized information and practical advice (Figure 2). Thus, it contributes to public education and awareness of environmental and waste management issues, improving community recycling behavior and practices [10].

The virtual assistant developed by 1MillionBot solves key waste management dilemmas, providing municipal authorities and citizens with invaluable support and addressing frequently asked questions about the amount of waste allowed, clarifications on dismantled furniture, and rules on leaving items on the street. The virtual assistant enables the city council to optimize its services and support all citizens by automating these consultations. This innovative approach aligns with the mission of Calvià 2000, a company dedicated to providing municipal services, including waste collection, urban space cleaning, water treatment, and management of the drinking water supply, in the Calvia municipality [19].

In conclusion, the AI solutions offered by 1MillionBot in terms of waste management are an innovative way to optimize processes and support environmental conservation efforts while providing tangible benefits to organizations and communities involved in waste management.

4.2. Litterbot

Litterbot is an advanced interactive system developed to help users identify items suitable for recycling. Its core mission is to simplify recycling by providing accurate and relevant information on local waste management regulations. The chatbot is based on sophisticated data analysis and natural language processing technologies, providing personalized and up-to-date answers in real time that are tailored to specific user requirements [20].

Litterbot’s mission is to simplify and facilitate the recycling process for people by providing accurate and relevant recycling information according to local regulations. Through technology, Litterbot aims to educate users about the environmental impact of recycling and encourage them to adopt more sustainable waste management practices.

What makes Litterbot so unique is its accessible and intuitive approach to providing recycling information. Using data analytics and natural language processing technologies, this chatbot offers fast, accurate answers tailored to users’ local needs and regulations. Typically, the chatbot asks the user to find out the material of which the product to be recycled is made, followed by a recycling number found on the product packaging. The purpose of these questions is to be able to advise the user whether they can recycle it or not (see Figure 3).

A distinctive aspect of Litterbot is its accessible and intuitive approach, which promotes understanding of the recycling process and encourages the adoption of sustainable waste management practices. Through its interactive communication, the system helps to raise awareness of the importance of recycling and promote more environmentally responsible behavior.

Implementing advanced technologies such as data analytics and natural language processing enables Litterbot to provide accurate and tailored responses in line with local requirements and regulations. Thus, Litterbot represents an innovative and efficient solution in waste management, helping to promote a recycling culture and protect the environment in an academic and accessible way.

In conclusion, Litterbot represents a significant innovative tool in waste management, providing a simple and effective solution to educate and engage people in recycling. Through its accessible approach and the advanced technologies that it uses, Litterbot contributes to promoting a recycling culture and protecting the environment.

4.3. GOA Plastic Waste Chatbot

Starting in June 2023, the first chatbot that allows users to report plastic waste in the public domain was launched and tested in Goa by the World Wildlife Fund (WWF)-India. It will be integrated into the WhatsApp application, allowing users to connect to the chatbot by this means. At first, it will ask some questions to understand the plastic management context: the type and quantity of plastic that is to be throw away. Also, it utilizes the location and users can send photos of the waste. The information collected by the chatbot will subsequently be analyzed by WWF-India to formulate a plan to address the problem areas, particularly focusing on the removal of waste sites in marine ecosystems. Following its trial phase, WWF-India intends to expand the deployment of chatbots to additional regions throughout the country.

Tourism operations in Goa and southern Maharashtra result in significant plastic waste entering the marine environment. These areas, renowned for their untouched natural landscapes, face the task of protecting endangered marine species and crucial habitats, such as coral reefs, from the threat posed by plastic pollution. WWF-India collaborated with Parmanoo Data Labs to develop the chatbot in response to this issue. Aditya Kakodkar, senior coordinator of marine biodiversity conservation at WWF-India’s Goa office, mentioned that “The project received funding from a financial services company as part of its CSR initiative.” Citizens in Goa can report plastic dumps by simply typing ‘plastic’ and sending it to WWF’s WhatsApp number.

The chatbot leads users through predetermined questions to collect essential information. “While there are apps designed to address similar issues, the chatbot eliminates the need for citizens to download an app, which could consume storage space on their phones”, explained the source. “The chatbot utilizes a widely used messaging service,” Kakodkar explained, and “Once the information is gathered through the chatbot, we will analyze it and collaborate with panchayats and manufacturers of disposable plastic products to develop a strategy to address the problem areas.” He also highlighted that the collected data will help map the scope of unmanaged plastic waste in natural environments and facilitate the identification of practical solutions. “The expansion of nature tourism, such as coral reef scuba diving and dolphin watching, has resulted in increased plastic pollution in these delicate ecosystems,” Kakodkar stated. “WWF-India is actively collaborating with dive and dolphin tour operators to enact efficient plastic waste management protocols. Recognizing the urgency of the intervention, we also introduced the chatbot, allowing citizens to report multiple dumping incidents in minutes”, he elaborated [21].

4.4. Waste Management Chatbot for the Environmental Organization—No Waste Ukraine

No Waste Ukraine, a non-profit environmental organization focused on social impact, required an automated aid to assist new individuals with waste sorting. Due to the underdeveloped municipal recycling system, insufficient recycling infrastructure such as bins and stations within the city, and ineffective sustainability initiatives coupled with flawed state regulations, cultivating a recycling culture in Kyiv was already quite a challenge. Implementing a digital assistant would streamline the process, reducing the workload and time invested by volunteers in answering inquiries through Telegram. Additionally, it would offer information on waste sorting with faster response times, improved flexibility, and easier maintenance. In March 2019, No Waste Ukraine and Infopulse unveiled the Waste Management Chatbot, which was immediately accessible on Telegram. Designed to offer guidance on waste sorting, this AI-powered chatbot project was crafted by students from NTUU KPI under the supervision of Infopulse professionals, with the following characteristics:

• An expanding database featuring categorized waste types and a map showcasing nearby recycling stations and bins that is continually being developed for user accessibility.
• The development of an educational chatbot using the Dialogflow development platform to facilitate question-and-answer interactions.
• The chatbot offers support in Ukrainian and Russian languages, ensuring accessibility for a broader user base.
• A convenient topic search.
• The incorporation of prompts, suggestions, and a compilation of helpful links that guide users through waste recycling techniques and offer guidance to reach the nearest recycling station.

The chatbot is designed to be user-friendly and intuitive, helping users understand the fundamentals of waste management, providing swift advice on waste sorting, locating the nearest waste disposal station, and offering information on its operating hours. It serves as a central hub for all queries related to waste recycling, significantly reducing response times and automating routine support tasks. This frees volunteers to focus on other crucial recycling program endeavors, ultimately fostering a more sustainable community. Accessible through Telegram, the waste sorting chatbot simplifies embracing eco-friendly practices (see Figure 4) [22].

As Yeugeniya Aratovskaya, the head of No Waste Ukraine, said, “We are very happy about the chatbot and eagerly await its launch, since each day we have to answer dozens of questions about waste sorting and environmental management. We hope that this bot will make information more accessible and useful for people while simplifying our daily work”.

4.5. SCC Chatbot

A council in the Midlands region was inundated with calls from residents asking about trash bin collections and waste management. Many of these questions could have been more varied, leading to a growing burden of time spent answering the same questions repeatedly. Collaborating with the council, SCC implemented an intelligent chatbot service that uses natural language understanding to convert caller speech into text, streamlining the process of addressing these inquiries efficiently.

The system uses AI to identify keywords or phrases and deliver automated voice responses based on predefined criteria. This functionality was seamlessly integrated into the existing call center infrastructure of the council. In cases where the chatbot cannot resolve queries, they are directed to a member of the council’s waste management team. Additionally, calls can be rerouted based on sentiment analysis; if the system detects increasing frustration from the caller, it automatically directs the call to a human agent, efficiently addressing the issue. Crucially, the chatbot can handle multiple calls simultaneously. This feature helps prevent the potential overwhelm of the waste management team in instances of sudden surges in public inquiries, ensuring smooth operations and effective allocation of resources. This straightforward application effectively uses natural language processing, AI, and analytics technology to streamline staff handling inquiries, ensuring consistent and efficient service delivery to the public. These advances are necessary as the only other viable options are to increase staffing levels, leading to higher costs, or to implement traditional marketing efforts or website enhancements, which may not guarantee impactful results. Although the council website already provides comprehensive information on garbage and recycling, a direct mailing and marketing campaign might have yielded little benefit. We can see in Figure 5 the SCC Chatbot logo [23].

In contrast, the chatbot enables the council to manage a higher volume of inquiries in a standardized and professional manner without substantial ongoing expenses. For more complex situations, callers can still be directed to staff members. This approach allows the team to allocate more time to addressing challenging or urgent issues rather than to handling routine public inquiries. Furthermore, the system can be seamlessly integrated into the council’s social media channels, enabling the implementation of a waste management chatbot on platforms such as Facebook to further reduce call center calls [23].

4.6. Chatbots’ Comparison

In this comparative examination, we evaluate and compare various recycling chatbots using a selection of fundamental benchmarks. These benchmarks encompass the technological functionalities, elements of the user interface, and security protocols crucial for gauging the efficiency and dependability of these chatbot systems. These criteria include the following:

• Natural Language Processing (NLP): the ability of the chatbot to understand and respond to user queries in natural language, ensuring efficient communication and user engagement.
• Image Processing Capabilities: the chatbot’s ability to process and interpret images related to recycling queries facilitates enhanced user interaction and problem-solving.
• Provision of Relevant Information: the capacity of the chatbot to provide accurate and relevant information about recycling practices, regulations, and environmental initiatives.
• Integrated Questionnaire: whether the chatbot incorporates a structured questionnaire to gather user preferences, habits, or requirements, thus personalizing the user experience and recommendations.
• Human Agent Redirection: the mechanism employed by the chatbot to redirect users to human agents or customer support representatives in cases where the query cannot be adequately addressed through automated responses.
• POI (Point of Interest) Location Assistance: the chatbot’s ability to assist users in locating relevant points of interest (POIs), such as recycling centers, drop-off points, or sustainable product vendors, using interactive maps or geolocation services.
• Security and Privacy: the measures implemented by the chatbot platform to ensure the security and privacy of user data, including data encryption, compliance with regulations (e.g., GDPR), and transparent data handling policies.

This comparative analysis (see

Table 1. Chatbots’ comparison.

	1MillionBot	LitterBot	SCC Chatbot	No Waste Ukraine	GOA Chatbot
NLP	✔	✔	✔	✔	✔
Platform integration	✔ (Multiple platforms)	❌	✔ (Call center service)	✔ (Telegram)	✔ (WhatsApp)
Image processing	❌	❌	❌	❌	✔
Information provider	✔	✔	✔	✔	✔
Questionnaire	✔	✔	❌	❌	❌
Human redirection	❌	❌	✔	❌	❌
Proximity recycling POI localization	❌	❌	❌	✔	❌
Security and privacy	✔	✔	✔	✔	✔

) aims to provide information on their strengths, limitations, and general suitability to promote recycling awareness and participation among users by systematically evaluating these criteria in different recycling chatbots. The necessary information was retrieved from the chatbots’ descriptions on their websites.

Following the comparative analysis of different recycling chatbots, it becomes evident that these platforms stand out through distinct approaches to addressing waste management challenges.

When it comes to natural language processing, information providing, and security and privacy, all of the chatbots showcase remarkable proficiency in NLP, ensuring accurate understanding and effective construction of responses to user queries. They also excel in providing up-to-date information across relevant topics and prioritize user security and privacy, safeguarding against malicious actions. Regarding the platforms that the chatbots can be integrated into, 1MillionBot is more versatile compared to the others, as it can be incorporated into multiple applications according to the user’s needs. SCC Chatbot, because it represents a local council initiative (Midlands), is integrated into a call center specific to the council, from where it can respond to citizens’ queries about waste management and rubbish bin collections, offloading some of the staff’s work to the chatbot. No Waste Ukraine and GOA Chatbot are integrated into popular messenger applications, making them available to as many people as possible (unlike SCC; those two are larger-scale projects). GOA Chatbot distinguishes itself with its robust image processing capabilities, enabling users to submit photos of plastic waste, thereby enriching the depth of the data collected. This feature sets it apart from other chatbots in terms of its data collection capabilities. With regard to questionnaire integration, 1MillionBot and LitterBot distinguish themselves by integrating questionnaires and providing a personalized experience to users. For example, LitterBot can decide whether a piece of rubbish can be recycled or not by asking the user some questions about the location they are in, what they want to recycle, and if the material they choose is recyclable, in which category it belongs based on a unique category number. As SCC helps the council of Midlands by automating the call center experience, when the robot does not have enough information about a subject or it is not trained sufficiently, it must have a fallback. This is why it has the capability to redirect to a human agent (a council member, in our case) who can further help the citizen in need. No Waste Ukraine excels in proximity recycling point-of-interest localization, leveraging geolocation and map integration to guide users to nearby recycling facilities. This feature underscores its commitment to environmental sustainability and user convenience. Despite their variances, all chatbots share a common emphasis on delivering precise and pertinent information regarding recycling practices, thus fostering community awareness and encouraging engagement in the adoption of more sustainable practices within the circular economy. In conclusion, the range of approaches and functionalities exhibited by these chatbots constitutes a noteworthy contribution to the promotion of sustainable waste management in the context of the circular economy.

5. Innovative Concepts Advancing the Circular Economy: Harnessing the Potential of Chatbots for Environmental Benefits

AI applications are starting to take shape in the circular economy, bringing significant environmental benefits that create a cleaner planet and propel the progress of humanity. The ongoing enhancement of chatbots within this nascent field is underpinned by a vigilant examination of both the intricacies of computer application development and the dynamics of user engagement. This requires a thorough evaluation of user feedback, the consequential community-wide effects of chatbot use, the establishment of user trust, positive implications across various operational domains facilitated by chatbot, and a judicious consideration of the economic implications related to the development and implementation of applications [24].

One key contribution in our present research is the attempt to provide valuable insights conducive to improving the efficacy of chatbots within the domain of circular economy development and implementation. To this end, the analysis centers on three innovative features that are poised to optimize the utility of chatbot solutions deployed across the broader spectrum of circular economy practices. These three features encompass the following (see Figure 6):

• Consumption and recycling tracking;
• Personalized advice;
• Education and awareness.

In the following subsections, we describe these proposed innovative features more in detail.

5.1. Tracking of Consumption and Recycling

At the level of a (smart) city, it is essential to track the consumption of products (what and how much is consumed) and their recycling. The process of becoming aware of the consumption of products, by which each person should buy only as much as he needs, is necessary and mandatory in this society due to the current excessive consumption. Most people buy products that they only sometimes use or buy much more than what is required for living. In this field, consumer education must focus on product quality, price, responsible use, rational purchase of products, and their recycling. During the holidays, most problems arise in relation to the excessive consumption of products. In this way, the population buys more products than they can consume, which leads to the throwing away of many purchases made initially, especially food [25].

A chatbot implemented in the circular economy can help the human city/community and the individual who chaotically consumes products. Based on certain graphs and reports, statistics can be gathered both at the city level and at the person level.

When we refer to the city level, we want to see the following:

• What is the quantity of products and the category of products consumed in a month of everyday shopping and a month of excess shopping;
• Which are the periods when more is consumed;
• What is the percentage of citizens in the community who follow correct recycling procedures to protect the environment;
• What are the users’ interests in the application;
• What is the trust of people in the chatbot;
• What effects does product recycling have on the community’s economy and the environment.

When referring to the person level, it is essential to consider the following:

• How the monthly consumption of products evolves;
• Offering personalized help (advice, recommendations, a shopping schedule, and making a shopping list that includes only the necessary products and following this list) to reduce the number of products that are unnecessarily consumed [26];
• Whether or not the person recycles products and how we can influence this;
• The person’s knowledge about the need to recycle products;
• The person’s awareness of the excessive consumption of products;
• What are the amounts of money spent on purchases that are not necessary;
• What can this money be invested in.

In accordance with the recommendations, we advocate the subsequent implementation methodology, articulated through a structured sequence of three fundamental steps (see Figure 7).

Step 1: Track the products that a person buys

It is necessary to create a database with a table for people, a table for products, and another table that shows the number of products consumed per month by a person. This can be achieved by identifying the person at the time of payment in the stores. A person can be identified by their bank account if they pay by card. After using the card and scanning all of the products, along with creating a tax receipt, a record can be made in the consumption table where the products purchased by the person and the quantities for each purchased product will be entered. If the person pays in cash, they can be identified by other means, such as requesting an SSN or identification by a store shopping card [27,28].

Step 2: Tracking recycled products

For this to be possible, the following is necessary:

• The city should have waste collection points;
• A program (established by the city hall in collaboration with the company that has a contract with the city hall for waste recycling) of waste collection (by waste category) should be known to all citizens of the city;
• The city should be equipped with intelligent dumpsters.

The intelligent dumpsters recognize products bought for recycling by different people, determine their quantities, and then record them in the database. Therefore, a monthly report can be prepared at the city level containing the quantities purchased, the recycled quantities, and the remaining quantities.

Convincing people to use waste collection points and actively participate in recycling initiatives requires an approach that considers different views and behaviors related to shopping and waste management. Here are some strategies that can be implemented:

-

Awareness and education

Providing accessible and easy-to-understand information about the benefits of recycling and its impact on the environment can raise awareness. Launching awareness campaigns that highlight the importance of waste collection points and their role in the circular economy can also contribute to increasing awareness and sensitization of the population.

-

Rewards and partnerships

Implementing a reward system for people who use collection points can stimulate participation in recycling. Rewards may include discounts and coupons for local stores. Collaborating with local merchants to provide benefits to those who actively participate in recycling is a strong incentive to encourage recycling.

-

Accessibility

Placing collection points in convenient and accessible locations, such as shopping centers, parks, and areas frequented regularly by the community, contributes to active participation in recycling. Waste collection should be simple and effortless for users, with clear instructions and easy waste separation methods.

-

Participation of the entire community

Creating a sense of belonging and responsibility through the direct involvement of the community in the management and maintenance of collection points is an effective method by which the entire community is involved in the recycling process. Organizing workshops and local events that educate and engage the community in sustainable recycling practices is another effective method.

-

Permanent feedback

Providing user feedback leads to an understanding of the barriers and challenges that the community faces in using collection points. The continuous improvement in the services and facilities offered at the collection points must be based on the suggestions and needs of the community.

If we want to make the same report at the personal level, we will first need to identify each person. Thus, an intelligent dumpster can locate the person in several ways, such as by asking them to create an account and then log into it to be able to recycle the respective products, sending a confirmation link by email, sending a code via SMS, using facial recognition, etc. Once the person is recognized, the number of products recycled by that person can be recorded in the database [29].

Step 3: Generate graphs and reports

Based on the information obtained from steps 1 and 2, different graphs can be generated to see the product consumption achieved. The graphs can be generated at the city level for all people, and based on them, some advice can be suggested to the city hall and circular economy bodies to help improve the quality of people’s lives and the environment. Graphs and reports can also be generated for a person through messaging applications such as email, Facebook, etc. Also, individual tips can be sent by the chatbot to the respective people [30].

A job that needs to be incorporated into the system can be defined as collecting the data at the beginning of each month for the previous month to create graphs and reports. It needs to take the data from the database and pass it to a service that deals with charts and reports [31]. After obtaining these documents, the chatbot can suggest different tips based on the information in the database.

The analysis can be performed to cover longer or shorter intervals, not just every month (once a year, every 3 months, or every 6 months).

This functionality would lead to the following results:

• An increased awareness of the importance of recycling products and materials;
• Helping people to understand when they have purchased more products than they need and how to control these momentary impulses, and directing their attention to other beneficial and enjoyable things;
• People will be aware of which products and in which quantities they bought in excess in order to avoid such situations in the future.

5.2. Personalized Counseling

5.2.1. What Is Personalized Counseling?

Personalized counseling is a multistep process. This modern approach puts the wealth of information gathered to use by technology. The information collected refers to the following:

• The buyer’s way of thinking and acting;
• The internal and external factors that cause the buyer to make more purchases, even if they do not necessarily need these products;
• Buyer preferences throughout the year, but especially during periods of excessive consumption;
• User behavior that is observed and analyzed over short or more extended periods.

These actions identify and provide users with advice, ideas, suggestions, and personalized recommendations. Chatbot functionality is based on advanced AI algorithms and data analysis. Thus, the chatbot creates a detailed profile for each user, which substantially helps to define, implement, and outline a circular lifestyle [32].

5.2.2. The Implementation of a Personalized Counseling Function for a Circular Lifestyle

The circular lifestyle is a new and modern approach that includes management, increased efficiency, sustainability, new opportunities, responsible choices, and assumed behaviors. In this process, nothing is thrown away. Everything is reused. The original traditional model, which recommended that the consumer buy the product, use it, and throw it away if it is used or no longer meets the criteria of choice and use, is replaced by a modern model with superior efficiency not only for the consumer and the field of activity of where the product comes from, but also for the areas of activity in which this product is reused and then recycled. The traditional model included actions such as taking, buying, producing, acquiring, using, and throwing away. By implementing a personalized counseling functionality for a circular lifestyle, opportunities are sought and identified through which users are guided to make responsible choices and behave in a way that supports this endeavor. In the medium and long term, the process also contributes to changing the mindset and lifestyle of all users [33].

The intelligence, creativity, and inventiveness of people have no limits. We see these amazing things all around us all of the time and marvel at discoveries and ingenious solutions that break down all of the barriers of imagination. The chatbot is an AI application that communicates with users by providing helpful information and real support in any field. The developers of these IT solutions have integrated a personalized advice functionality, creating the opportunity for chatbots to take on the role of personal guides in the process of transforming consumer behavior and adopting a circular lifestyle.

The chatbot’s personalized advisor role involves the following five procedures:

1.

Detailed analysis of preferences and behaviors

The chatbot collects information about user preferences, internal and external factors that lead them to uncontrollably buy products, peak consumption periods, and user behaviors over short, medium, and extended periods. Sets of questions are established with which the chatbot identifies problems, quickly collects data, and outlines personalized recommendations and advice for each discovered problem. Interactions with the chatbot are customized on the basis of the user’s needs and the identified problem. On the basis of all of the data obtained, the chatbot builds a detailed profile for each user. This profile is constantly improved and modified, depending on the data that the chatbot collects about the user [34,35].

2.

Creation of user profile

The chatbot creates the user’s personalized profile based on the collected data. This profile includes information such as consumption preferences, the amount of products purchased, periods of excess shopping, normal shopping periods, and products purchased during both of these periods, purchasing habits, their level of awareness of environmental issues, the amount of money spent on shopping to excess, what solutions are identified to promote the spending this money only for applicable purchases or for the profitable investment of this money, what solutions are recommended for adopting a circular lifestyle, and what are the short-, medium-, and long-term impacts on the user of addressing their responsibility for their lifestyle [36,37].

3.

Personalized recommendations and continuing education

Using advanced algorithms, the chatbot generates personalized recommendations. These may include suggestions such as the following [38]:

• Healthy products, not chemically treated ones (food).
• Products made in nearby local communities, where the consumer can go to pick/purchase the fresh product right from the seller’s household or garden.
• Durable products that can be used for a longer period.
• Recycling options for each product category.
• Waste reduction.

In parallel, the chatbot provides personalized educational information to increase users’ awareness of the impact of their decisions on the environment [39].

4.

Facilitating sustainable decision making

The chatbot offers recommendations to each user but also facilitates the process of making sustainable decisions by maintaining ongoing collaboration, communicating interactively, and asking personal questions of the user. Depending on the consumer’s evolution toward a circular lifestyle, the questions are kept as they were originally asked or are modified to meet the needs of each user. The chatbot helps the user better understand what options they have and what changes to make and helps them make responsible choices [40].

An important part of the functionality of personalized advice for a circular lifestyle is the continuous education of users. This involves the following:

• Provision by the chatbot of information regarding the impact of consumer decisions on the environment.
• Offering sustainable options.

The better-informed users are, the more sustainable decisions they will make and the more responsible they will be for themselves and the environment [41].

5.

Measuring progress and rewarding users

To keep users engaged and willing to change, the chatbot can provide continuous feedback on their short-, medium-, and long-term progress. In the context of the chatbot’s encouraging and supportive role, such actions may manifest in the form of positive reinforcement, expressed through praise or congratulations for each effective user activity. Additionally, the provision of rewards serves as a motivational mechanism to incentivize the user toward the implementation of novel ideas. The chatbot diligently assesses the ongoing impact of each user’s decisions in the pursuit of adopting a more circular lifestyle, thus contributing to a sustained and dynamic evaluation of user engagement with sustainable practices [42].

5.3. Education and Awareness

To encourage and support the adoption of a circular lifestyle, it is essential to educate and raise awareness of the benefits and positive impact of this approach. The chatbot is an innovative and interactive means of providing this information, stimulating dialogue, creating curiosity and interest, and educating users about the conscious and responsible approach to a circular lifestyle [43,44].

5.3.1. Chatbot Features for Education and Awareness

1.

Presentation of basic concepts

The chatbot can explain key concepts of the circular economy. These are fundamental concepts that help the users to understand the information in this field. Here are some examples: regeneration, reuse, recycling, resources, recovery, and restoration. The chatbot presents these keywords often used in the circular economy in a more accessible language that can also be understood by people who do not work in this field, do not have a rich knowledge base, or have not graduated from higher forms of training and specialization [45].

2.

Information on sustainable practices

In the dialogue with users, the chatbot presents information on habits and practices that people can adopt to behave responsibly and correctly. Situations like “So YES” and “So NO” are also exemplified to increase the impact and encourage people to adopt a circular lifestyle daily [46,47].

3.

Statistics and case studies

This functionality of the chatbot requires storage in its database of information on the circular economy that can be presented to users in the form of case studies and statistics that come with solid arguments supporting the approach of a circular lifestyle. Information can be presented by comparing traditional and modern circular models [48].

4.

Personalized questions and answers

The chatbot can answer personalized user questions by providing answers based on each person’s needs and challenges. Chatbot responses are constantly being improved. The chatbot also increases its database through dialogue with users. The solution identified to a problem of one user may be different to the case of another user raising the same problem. The information is tailored to the specific needs of each user [49].

5.

News about news

The app is constantly being improved to attract as many users as possible, increase trust in the chatbot, and provide new, verified, correct, and up-to-date circular economy information. The app provides alerts and updated news about news in the field, about developments and new implementations of AI in the circular economy, and about countries/communities/firms/people who have implemented or improved original solutions and thus are becoming role models and a true example of responsible behavior [50].

5.3.2. Benefits of Using Chatbot for Education and Awareness

• Accessibility

The application can be used at any time and from anywhere. The user only needs to be connected to the Internet. This is a significant opportunity that facilitates continuous learning and contributes to educating society about a healthy lifestyle, promoting a responsible approach, and promoting the circular economy [51].

2.

Interaction

Through dialogue with users, the chatbot provides an interactive experience, prompts questions, and provides personalized responses. All of this strengthens user understanding, contributing to awareness of the process and easier adoption of positive change [52].

3.

Personalization

The chatbot provides personalized information to each user. The answers provided and the sets of questions asked according to the need of each person lead to the following [53]:

• Increasing the number of users using the application;
• Increasing the commitment of the person involved in the change;
• Increasing the relevance of the information provided,
• Increasing user confidence in the chatbot.
• Ease of use

The application is easy to use and does not require complex skills such as being an IT specialist. The chatbot’s intuitive and conversational interface makes it easy to access information without requiring technical expertise. These elements attract an increasing number of users interested in positive changes in their lives and the lives of their loved ones [54].

5.4. Limitations of These Characteristics

Implementing a circular economy chatbot to track consumption and recycling, provide personalized advice, and contribute to education and awareness presents several limitations specific to technology, data access, and human interaction:

• Access to data and its accuracy

A chatbot depends on access to up-to-date and accurate databases to track consumption and recycling. Inaccurate or incomplete data can lead to erroneous recommendations that affect the quality of personalized advice and information provided.

2.

Customizing the chatbot

Customizing responses to user needs may be limited by the algorithm’s ability to understand and process complex variations in user behavior and preferences. Chatbots may have difficulty understanding the context or nuances of human language. Thus, inadequate answers or misunderstandings can be obtained.

3.

Education and awareness

A chatbot’s ability to educate and raise awareness is limited by how the information is presented and the user’s ability to engage in a meaningful conversation with a chatbot. Effective education and raising awareness also require a human factor, meaning discussions, debates, and interpretations. These are difficult to achieve with a chatbot.

4.

Technology and security

Some users have limited access to technology. At the same time, users with limited digital skills may encounter difficulties in interacting with the chatbot. The implementation must ensure the protection of user data. This is a challenge in the safe management and storage of sensitive information.

5.

Sustainability of the solution

The development and permanent update of a chatbot require significant financial and human resources. The ability of the chatbot to adapt to the permanent changes that occur in the circular economy sector, the different behavior of consumers, and the technologies used in the development and implementation of the chatbot can be limited and constantly face new challenges.

6. Discussion and Conclusions

In this paper, we explore the potential of leveraging AI solutions, such as the chatbot, in fostering the implementation and development of a circular economic system. The investigation approaches this topic from various perspectives to gauge its effectiveness. As the main contributions of this paper, we achieve the following:

• Present a brief review of the literature on the integration of AI technologies, particularly chatbots, as a promising solution to address waste management challenges in the context of the circular economy;
• Systematically analyze and evaluate five representative examples from the existing circular economy chatbots, providing a comparison table between them and highlighting their similarities and what makes them unique in the context mentioned above;
• Propose some new ideas that can be implemented in such chatbots in order to improve their functionalities. Among these proposals, we mention the possibility of integrated product tracking (and all of the phases it goes through, including consumption and recycling) that can be shown through different generated reports, as well as graphical representations, and educating and making the user aware by providing targeted information, displaying the latest news in the field, and using some aspects of gamification (the application of game-design elements and principles in non-game contexts to engage users and solve problems) to reward them if their decisions are aligned with the circular principles;
• Mention, as for the benefits, some ideas which can improve accessibility (the developed solutions are accessible anytime and anywhere, requiring only an internet connection), promote user interaction (the chatbot offers an interactive encounter, initiates inquiries, and delivers tailored responses), and keep users engaged through positive feedback;
• Identify some limitations regarding the possible efficiency of the innovative ideas proposed as measures to implement better recycling chatbots in the future.

In conclusion, we strongly believe in the increasingly significant role that AI will play in the evolution of IT applications targeting smart city infrastructures. In addition, it is anticipated that a proliferation of dedicated applications will emerge, specifically engineered to optimize recycling efficiency and improve waste management within urban settings.

In our future research directions, we intend to focus on the practical application of the inherent advantages of chatbot programs. This effort will be directed toward the actual development of a chatbot integrated in the circular economy paradigm, specifically designed to improve waste management initiatives in the context of a smart city.