The University of Baltimore was host to Cyber Discovery Camp from 2013 to 2015. This camp was a week-long experience for rising high school sophomores that covered a number of topics relating to “cyber”. One of the most exciting parts of the week was the integration of robotics into the camp. We used Boe-Bots from Parallax and students needed to code the robots to go through a maze, launch a golf ball, play soccer, and compete in a battle arena collecting marbles via remote control.
Boe-bots cost about $180USD and our camp had 20. After years of abuse, the bots started to deteriorate and we needed to replace the processor parts. A replacement Boe-board (Board of Education) is around $100USD. Unfortunately, our funding deteriorated and we had to cancel the 2016 camp. In 2017, we received a small amount of funding from the University of Baltimore Foundation to run a one-day camp. We needed to update our aging tech within a small budget to make this camp work.
In order to do this, we looked at what we had and what we needed. First, the chassis and the servos of the Boe-bots were still working well. The Parallax processor and associated board was the problem. We knew we needed to build a new board to put on the bots that was rugged and cheap. Since the curriculum was first designed, there have a been a number of significant advances in low-cost microprocessor technologies and we were excited in exploring the use of Arduinos in our bots.
We evaluated different Arduino boards (official and unofficial) and found a WeMos D1R2 which had the same shape of the Arduino board but had wi-fi built in. By using wi-fi, we could simplify our activities that relied on remote control. We had access to 3D printers so we decided to build our own shapes.
We started designing a physical board with cardboard and then moved to Tinkercad. There were multiple iterations (as any good design should have!) and we eventually settled on one that used the same posts as the Boe-boards but were slightly larger to accommodate handles.
We added small, adhesive breadboards to the open space in the front of the body. This enabled us to use the existing servo hook ups with the microcontrollers by using male-to-male wires and pave the way for any future upgrades.
One of the challenges for the camp was to play marble soccer with the robots. We had been using custom designed metal jaws that were fabricated by a previous partner. We don’t have the capability to do any metal fabrication (and I thought they were overkill anyway), so, I designed up some front scoops. I was inspired by the mBot and made up a scoop with a face.
We spent a week printing our final pieces and had the camp faculty make an assembly line to build the WhimsyBots from the former Boe-bots.
In the end, we had a few problems. First, it takes a long time to 3D print things. We used an *old* MakerBot Replicator Dual for a majority of them because it could still print faster than our our new Monoprice 3D printers. Of course, 3D printing is more voodoo than science and the MakerBot eventually stopped working significantly slowing down our production.
Second, cheap electronics are just that, cheap. We had some problems with the wires and breadboards not making connections. Our higher quality jumper wires worked well while the discount ones from Amazon were hit or miss. We had built eight WhimsyBots and ended up with only two working ones at the end of the day due to broken electronic components. Adding to the troubles, using default wifi settings on these boards created a cacophony of radio interference on the same channel and sometimes remote control messages weren’t getting across.
We think the WhimsyBots are a good retrofit to Boe-bots or a great starting place for designing robots for education from scratch. Here are the resource files so you can replicate what we’ve done:
GitHub Repository (still a work in progress)
If you build any WhimsyBots, please share with me at Greg Walsh on Twitter!