This is the second day of the symposium and I thought I'd write some observations.
First, there are more than 65 people attending this symposium. AAAI had to give us the largest room available! Obviously there is a lot of interest in educational robotics. Thanks to Doug Blank and Zach Dodds for organizing a great set of presentations, discussions and demos.
Yesterday, we split up into ad hoc groups to discuss "Top 10 things NOT to do with robots in education". Here are some highlights:
1. Do NOT drop the robot. Ensure that they do not roll off the table.
2. Do NOT attempt to get into robots if you do not have sufficient financial and personnel resources. I.e. make sure you have adequate support from your institution.
3. Do NOT get robots to go too fast.
4. Do NOT allow last minute changes in robot demos/competitions.
5. Do NOT assume that all outcomes will be met for any given robot assignment.
6. Do NOT try to be too efficient in your course/syllabus. Allow time for discussing unexpected things.
7. Do NOT put too much emphasis on the final grade.
8. Do NOT limit exercises to scripted tasks. Allow for creativity and open endedness.
9. Do NOT lock out lab access. I.e. allow access at all times.
10. Do NOT spend too much time on hardware debugging.
Not necessarily in any particular order, but they do shed light on the issues and needs in the use of robots in the class room. IPRE's reason for existence is to partly attempt to overcome these in a productive and meaningful manner. Here are some of our goals:
1. Make robots affordable and personal. Every student owns their robot and has total access to it at all times
2. Make them rugged.
3. Let the curriculum drive the design of the robot. The course materials we are developing will be tied into the robot so that all exercises and materials have sound pedagogical goals and will engage students appropriately.
4. Students will not have any hardware issues to deal with.
Fred Martin from U-MASS Lowell gave a really good paper titled, "Real Robots Don't Drive Straight". An instant classic, IMHO. Fred makes two important points: (1) Feedback is an essential component of robot (behavior) design and robots can and should make every attempt to expose students to the concept of feedback and its central role in engineering. (2) As the title suggests, robots do not go straight and that should be an expectation. Students have prior misconceptions about robots that they are perfect mechanical things that will go in a straight line and will always turn the exact number of degrees requested. Fred cited that he was dissappointed that the LEGO Mindstorms NXT now includes motors that can guarantee straight line motion, though it was pointed out that you do not have to use that feature. Fred gave nice exmaples of trying to swim in a straight line or teaching a kid to ride bicycle. Without any feedback, it is nearly impossible for humans to swim or ride a bike in a straight line (hence the criticality of feedback).
There are of course applications of robots where straight line or precise motion is needed. But in most educational situations, the robots are being used as creatures rather than mechanisms for precise motions. This also directly leads to implications on the kinds of exercises one should deisgn for students.