Bachelor of Science (Honours) in Design and Innovation

AI-Powered Recycling Assistant: Improving Household Recycling Rates Through Conversational Guidance
A case study on the dissertation presented for my Bachelor of Science (Honours) in Design and Innovation

Project Overview

The growing problem of household recycling contamination has significant environmental consequences, contributing to landfill waste and hindering recycling efforts. To address this issue, I developed a concept for an AI-powered recycling assistant (ART) designed to provide real-time sorting guidance and personalised user education.

Problem Statement

Many individuals struggle with understanding complex recycling rules and regulations, leading to confusion and unintentional contamination. This lack of knowledge hinders recycling efficiency and negatively impacts environmental sustainability.

To delve into the causes of recycling failure, an expert interview was conducted with Ewan Hill, Operations Director at Levenseat, a waste management company. He identified that a lack of household knowledge and care impacted the recycling success rates. During a site visit, he showcased how quantities of waste had to be removed from the process early due to contamination (removed waste becomes unrecyclable), as contaminated products could damage sorting equipment.

The Waste & Resources Action Programme (WRAP) provided additional sources of information through reports (WRAP, 2024) and case studies (WRAP, 2021) highlighting the scale of household contamination.

Figure 2 shows the number of households contributing(84%)
Figure 4 highlights the high quantity of contaminated items being disposed of incorrectly

Vision

The vision prioritises ease and simplicity, recognising that overly complex recycling systems create a significant barrier to correct disposal, especially for individuals with limited time, mobility, or prior knowledge.  The vision emphasises minimising contamination to directly support maximising resource recovery, aligning with a strong commitment to sustainability. This concept was inspired by insights from Ewan Hill, Operations Director at Levenseat, who likened recyclable materials within waste streams to precious resources awaiting extraction.

The overall vision of this project was to bridge the gap between recycling knowledge and action by providing a tailored, interactive platform that inspires and motivates households to adopt eco-friendly behaviours.

Research and Analysis

Several research methods, from the module's project toolkit on research skills, were employed to gain a comprehensive understanding of the recycling context. Two methods included:

Review of Academic Literature: researching relevant academic papers, such as:

McDonald & Oates's exploration of reasons for non-compliance with recycling schemes provided valuable insights including a catalogue of participants' responses (Appendix G). (McDonald and Oates, 2003)
Keramitsoglou & Tsagarakis's survey highlighted a disconnect between a willingness to recycle (91.8%) and a lack of knowledge of proper disposal (73.2%). (Keramitsoglou and Tsagarakis, 2013)

Expert Interview: Conduct 2 semi-structured interviews with Ewan Hill, a recycling sector specialist. These interviews provided first-hand insights into common misconceptions about recycling and the importance of user-friendly design in promoting effective recycling.

People

A Mendelow stakeholder matrix identified households as the primary project stakeholders and categorised them into Elderly Households, General Households, and Households with Disabilities. Understanding their unique needs was essential for creating an inclusive recycling solution.

To stay ahead of the curve, I conducted research into alternative recycling apps. My findings revealed several promising tools with a range of features. However, reviews were often mixed, stating problems with waste recognition accuracy, missing features and paywalls.

These did help highlight areas for improvements that the ART bot being conceptually designed could address.

Design Process

Ideas and options

Eight initial concepts were developed and evaluated using tools from the project toolkit (The Open University, 2024). Public feedback helped narrow the selection to three, including an A.I. product (initially the product was envisioned as a bin with built-in AI capabilities) and these alternatives:

Pneumatic waste collection: A system of underground pipes transporting household waste to sorting centres to streamline the collection and improve sorting efficiency.
RDF-tagged bin bags: Bags linked to households, with contamination tracked and repeat offenders potentially facing warnings or fines.

Alternative A.I designs

Material Aspects

The product's design prioritises sustainability; wherever possible, researched alternatives replace conventional materials with a higher environmental impact. Plastics and precious minerals present challenges for maintaining a low environmental footprint.

Thermosetting polymers are needed for structural components but raise concerns due to their non-recyclable nature. A paper on sustainable alternatives to thermoset composites (Patricia Ares Elejoste et al., 2023) highlighted potential substitutes like bio-based or Furan resin which offers good thermal stability, no by-products during the curing reaction and are derived from renewable sources.

Solutions and Outcomes

The ART concept features the following key functionalities:

Visual Recognition: ART utilises advanced image processing algorithms to identify recyclable materials in real time.
Conversational Interface: Users can interact with ART through a natural language interface, asking questions and receiving personalised guidance.
Cloud-Hosted Database: ART accesses a cloud-based database of recycling information, ensuring up-to-date and accurate guidance.

By leveraging these features, ART empowers users to make informed recycling decisions and develop sustainable habits. It has the potential to significantly improve household recycling rates and contribute to a more environmentally friendly future.

Lessons Learned

Throughout the design process, I learned the importance of iterative testing and user feedback. By incorporating user insights, I was able to refine the ART concept to meet their specific needs and preferences. Additionally, the project highlighted the value of AI-powered solutions in addressing complex environmental challenges.  The AI-Powered Recycling Assistant represents a promising solution to the problem of household recycling contamination. By providing real-time guidance and personalized education, ART can empower users to make informed choices and contribute to a more sustainable future.

Bachelor of Science (Honours) in Design and Innovation

Project Overview

Problem Statement

Many individuals struggle with understanding complex recycling rules and regulations, leading to confusion and unintentional contamination. This lack of knowledge hinders recycling efficiency and negatively impacts environmental sustainability.

Figure 2 shows the number of households contributing(84%)

Figure 4 highlights the high quantity of contaminated items being disposed of incorrectly

Vision

The overall vision of this project was to bridge the gap between recycling knowledge and action by providing a tailored, interactive platform that inspires and motivates households to adopt eco-friendly behaviours.

Research and Analysis

Several research methods, from the module's project toolkit on research skills, were employed to gain a comprehensive understanding of the recycling context. Two methods included: Review of Academic Literature: researching relevant academic papers, such as:

McDonald & Oates's exploration of reasons for non-compliance with recycling schemes provided valuable insights including a catalogue of participants' responses (Appendix G). (McDonald and Oates, 2003)

Keramitsoglou & Tsagarakis's survey highlighted a disconnect between a willingness to recycle (91.8%) and a lack of knowledge of proper disposal (73.2%). (Keramitsoglou and Tsagarakis, 2013)