Concept
Erglove is an instant and long-term wrist posture feedback system that allows hotel housecleaners to gain awareness of their posture and change their behavior as they wipe. I worked with Bill Xu and Amil Vira on this project in Spring 2019 for DEA 4700: Applied Ergonomic Methods. The resulting design research paper was published and presented at the IASDR 2019 conference.
Prompt
Design an interactive product that informs a hotel worker of a current or potential ergonomic issue and encourages her/him to resolve the issue.
Process
We chose to investigate the experience of housecleaners specifically because they have the highest rates of injury among all hotel workers.
Early on, our group generated a wealth of ideas related to the problem space. We wanted to choose an idea that would fit the timeframe and abilities of our team members to complete. A wrist posture sensor came up as a potential exciting idea.
Next we conducted an expert interview at a local hotel to learn about the challenges hotel housecleaners face. We also got to observe them using a few tools to clean the hotel room.
Above is a summary of the main findings from our expert interview.
One key insight we learned is that this hotel attracts personnel from around the country and globe, and many cleaners learned improper techniques when they were younger, and thus need to relearn safer, ergonomic methods.
This led us to the question...
How might we enable cleaners to learn correct cleaning techniques?
How might we deliver flex sensor feedback to hotel cleaners to enable them to change their posture?
We used the Behavior Change Strategy Cards from Artefact Group to guide us. We narrowed the list of strategies down to the methods that made most sense for a wrist posture feedback glove:
1. Help the user make a commitment in advance
2. Provide immediate and ongoing feedback
Prototype
The Erglove prototype was made using a glove embedded with flex sensors, a micro-computer, and sewn-in LED lights. The flex sensors collect movement input from the worker using the glove and the LED lights would inform the workers about their wrist posture.
User Testing
To test the proof of concept, we conducted a pilot user study with using a Wizard-of-Oz approach.
To simulate the effect of the Erglove system, we asked participants to set a goal for how well they'd like to clean before they started the task (mimicking goal setting with a manager), and used a remote controlled light in the bottom left of the window to give feedback on wrist posture (mimicking feedback lights on the glove).
Our study focused on assessing the effectiveness of different types of feedback on participants’ experience and perception of our design. One of our success criteria was an improvement in wrist posture between an initial and final round of cleaning. We counted the number of wrist posture mistakes made while cleaning. The other criterion was a positive perception towards Erglove. This was measured through a post session survey.
There were two conditions in our study: one group performed the task using our glove with real-time feedback (via light), plus post task feedback (via verbal reporting); the second group performed the same task using our glove, but with post task feedback only.
The experimental group was more positive about the feedback and the system as a whole compared to the control group. All participants (both experimental and control groups) saw improvement in their cleaning technique between the first and second rounds, while the experimental group with real-time feedback showed higher perceived system usefulness. In addition, as observed from our expert interviews, ergonomic instruction is critical in affecting behavior change. Our user study further demonstrated the importance and impact of feedback and usage instruction.
Reflection
This project was a great chance to dive deep into a real life ergonomic issue that affects thousands of workers, and was a good exercise in rapid prototyping - our group had around 8 weeks to pinpoint the problem, conduct interviews, develop a prototype, and test the product with users.
Final Video