Introducing an innovative concept
In working towards an innovative community-based social system (as described in Social Recipes), I designed and evaluated another component that could be part of this system. This component or concept is E-COmate.
E-COmate is an augmented bin that captures and visualizes domestic food waste data and feeds this back to users for more readily comprehensible and accessible information.
E-COmate shows consumers their food waste amounts within their own kitchen environment with the intention to elicit reflections on a daily basis without the requirement of conscious effort. E-COmate uses metaphors to visualize food waste amounts that can be directly linked to consumers’ daily concerns, in this case, wasted servings. E-COmate also includes social comparison information to provide indications of waste amounts relatively to others. This was done with the aim at engaging users in social interaction (e.g., collaboration and competition).
The aim of this work was to understand how to visualize food waste information to foster sustainable behavior in domestic environments. Specifically,
I was interested to explore how eco-feedback (or E-COmate) impacts awareness, active engagements, and how it should be designed to motivate people in preventing food waste.
A deployment study in a student residence
To address this question, I designed and developed E-COmate, evaluated its impact in a 2-month deployment in a student residence, and designed and evaluated additional display visualizations. During the deployment, I gathered data through visual inspections, a questionnaire, and semi-structured interviews. Findings showed that visualizing food waste amounts helped in enhancing awareness on moments of disposal, active engagements in the grocery store, and hence in reducing waste of edible or previously edible foods (i.e., perishables). I also found an impact of social comparison information on participants’ level of motivation to reduce waste. Based on these findings, I provided design implications on what and how to visualize food waste data: for example, specificity, costs related to wasted foods, and social competition were considered main motivators.
Tools, skills and experience
As the lead of this PhD project, I defined the research goals and system requirements, designed the experiment, collected data, and performed the analysis. I collaborated with a student assistant in electrical engineering and a computer scientist for the development.
- Defined system requirements for the Raspberry Pi
- Prepared and embedded the prototype in physical bins: used Illustrator to make the template for laser cutting
- UI Design and development, specifications for server-side processing, and dealt with SQL and PHP
- Prepared procedures and protocols for the user studies
- Performed usability tests
- Recruited participants
- Deployed and collected data in a real-world setting (longitudinal)
- Conducted systematically visual inspections (waste audits)
- Designed and employed surveys with Likert scales and open-ended questions
- Designed and conducted retrospective interviews
- Conducted observational research
- Explored quantitative data using Tableau
- Applied qualitative analysis: coding, inter-rater reliability, thematic analysis and triangulation
The main challenges of this work
The main challenges were technological. The prototype was supposed to stay on 24/7, but Wi-FI interruptions and daily trash bag replacements by Janitors were factors that could crash the system. It needed regular rebooting. These limitation had impacts on data collection and analysis.
Another challenge was the management and involvement of stakeholders (the university, the student housing, ethics committee, the carpenter, the janitors, waste-related facilities, and IT facility). Different individuals were involved in many different tasks to get the user study started within the rules and ethics of the University.
The findings and implications are expected to serve as a basis for understanding basic requirements for effective eco-feedback design within the context of food waste and sustainability, a topic that has not received full attention yet in HCI.
This work is submitted to the Proceedings of ACM Conference on Computer-Human Interaction: Lim, V., Bartram, L., Funk, M., Rauterberg, M. (2017). To eat or not to eat: an evaluation of the impacts of eco-feedback in a student residence.