April 15, 2019
Social Robot Wrangling: Assembling Tacoma’s First Social Robot
Blog post written by Sally Ho
Our team was excited to get our first EMAR V4 on the UWT campus in Winter quarter. As the Social Robot Wrangler, my goal this quarter was to assemble EMAR and improve the instructions that came with it to help teams assemble it in the future. Our team is excited to have EMAR at UWT because we can share it with our community partners, gather more data about EMAR, and explore how we can continue improving it over time.
Trouble getting started
The process of putting EMAR together began when I was handed two tablets and two felt boxes, one that was bigger and one that was smaller. The instructions said that the two tablets would be different sizes, but I noticed both of the tablets I received were the same size. As a result, the face tablet did not fit inside the head.
I also noticed that both tablets came factory reset, and I was not sure what to do so I left the information blank and set up the tablets with no information attached. One of the tablets also came dead on arrival. In addition, there was no belly URL for the body of the robot. When I turned the tablets on, I could only see the face.
To address these issues, I reached out to the Seattle team with my questions. The team worked on getting the stress slider that would appear on EMAR’s belly to work and I was able to start putting EMAR together.
Putting EMAR together
Once I had spoken with the Seattle team, I began building EMAR. I started by measuring and cutting a panel on both of the boxes to fit the tablets into.
Figure 1: Drawing the panel outline for EMAR’s face
I could not find a way to fit the face tablet in comfortably no matter how much I forced the tablet to go in. To fix this issue, I cut slits along the edge of EMAR’s head so the tablet would easily fit inside.
Figure 2: EMAR’s face tablet hangs slightly over the edge of the box, making it look like a pair of ears
Once I was finished assembling EMAR’s head, I went on to work on assembling its body. This was a little bit trickier because there was nothing holding up the tablet. I decided to focus first on getting the panel cut out, and worry about holding up the tablet later.
Figure 3: The belly tablet initially displayed a solid yellow screen
Once I had the basic structure working, I was able to focus on making the tablet for the body stay up. I discovered that if I put enough pressure on the tablet against the cut out panels, it would stay up by itself. I then took the second part of the box and rested the tablet against the surface so it would create enough pressure to hold it. Until I have an alternative method for keeping the body tablet up and positioned, this will have to do.
Figure 4: The felt pieces on the backside of EMAR hold the tablets in place
When I was finished putting EMAR together, our team was able to take EMAR into the field and share it with the Girls Who Code and local high school teachers.
Figure 5: The Tacoma Team’s final working prototype of EMAR V4
Revising the instructions
In addition to assembling EMAR, I updated the instructions for EMAR to make assembling it easier for future robot wranglers. When I began trying to assemble EMAR, it was confusing to read the instructions because some of the materials were missing or out-of-date and the information was incomplete. To help future teams, I added instructions to cut slits on the side of the head panels so the face tablet will fit into the box easier because that was something I had trouble implementing.
Furthermore, the original instructions had stated that there would be two tablets, one that would be smaller for the face and one that would be bigger for the body. However, both of the tablets were the same size, and the face tablet didn’t fit inside the head. I rewrote the instructions to include that the tablets would be the same size. I also added a section detailing how to cut two slits in the head box so the face tablet would fit, making the V4 look like it has ears.
Moving forward
Overall, I learned a lot from assembling EMAR. It was frustrating to assemble EMAR when the tablet wouldn’t fit in the box, so I learned to communicate with the team to make sure that it was okay to cut slits in the sides of the box. I also had to communicate with the Seattle team to see how their development was because we needed them to update EMAR so we could use them for field studies. A lot of work went into making this prototype and even though I only recently joined, I’m very excited to see how the future prototypes will turn out!
Next quarter, we will be continuing to work toward improving the instructions for assembling EMAR. We are looking forward to sharing V4 with more community members in the future so we can continue building our understanding of how a social robot can address teen stress.
