Portolio of David Elkan-Gonzalez
A brief video of the vecna cares event. k9 shows up at 3:23

Here k9 is driving through the crowd as we get ready for our heat at Vecna Cares Robot competition

K9 Construction
Team Size: 4
Project Duration: 4 weeks
Year Completed: Senior Year
Key Skills: Mechanical Design, Computer Vision, Robot Nav, Circuits
Vecna Technologies, a software company that has recently secured a foothold in robotics, has begun hosting an annual robotics race. The event is more about people coming together to show off robots and get kids excited about learning than about actual competition. Determined to make an entry to this event, I got some friends together and with a paltry budget and only a few weeks we set out to recreate the greatest robot ever conceived; Doctor Who's robot companion, K-9. Once completed our robotic K-9 would traverse a short course, pick up a cup, and return to the starting line. Oh and one thing I forgot to mention. Most of the entrants to this event were professional robotics companies, including Vecna, IRobot, or post-graduate robotics labs, including ones from MIT, and WPI. Their entries would be the products of years of work and many thousands or even millions of dollars. This particular detail was picked up on by several news groups including Canadian Discovery Channel, who profiled and followed our team through the k-9 creation process.
With a highly limited schedule and budget, we set out with little but ambition to fuel us. We scavenged nearly all of the components and spent several sleepless nights getting everything together. Our processor was a Raspberry Pi. We connected up a camera which would be used for detecting the cup. A simple grabber was rigged up with some laser cut plaster and a servo. We even acquired a spare neato lidar for localization and navigation. We even acquired some spare wheel-chair motors a friend had lying around. We had many ambitious plans, and it would be somewhat of an exaggeration to claim we fully succeeded. For a number of reasons our lidar did not end up working correctly, so without navigation sensors we ended up remote controlling our robot. In fact pretty much everyone ended up doing this; it turns out hay bails in direct sunlight make lidar very hard to read. Also our grabber got lodged under the robot en route and tore off. But that
was fine because our back-up grabber (our smallest team member who literally road K-9) worked just fine. Despite the lack of total success, we were very proud to have created a pretty awesome robot, and I had an awesome time putting together and leading this team.