Subscribe via Email

Friday, March 25, 2016

Today in the AA [#2]: TEDx - What Can We Learn From Robot Athletes?

I vowed a while back to continue blogging, and here I am. Over the past couple months, a lot has taken place in the Autonomous Agents lab here at the University of Manitoba, and with some very exciting things (in which I'll mention later on this blog) in the future, I think it's important to track more of it here.

Anyway, let's talk about ted talks!




Jacky Baltes, Professor and Co-Director of the lab, gave a talk in regards to the importance of robotics competitions for both research, and skill development for students.
The skill development part was spot on: through the last couple days I've had quite a fun time going from my work on Arash (the Iranian teen-size robot) to the Darwin OP2 in which would be performing on stage, live, during this talk.


A rehearsal took place the night before, with the following day being last minute checkups. Things were "working", whatever that means in robotics. What once was working is now breakdancing face-first in front of a live audience...
At least, that was my fear. 


 




We arrived the next day for the 5pm talk a modest 2 and a half hours early, quickly setting up base during the brief 15-minute break intervals between speakers. You can begin to feel the tension as the time passes, I noticed this around the same time I claimed a  good half of the snack table.

Can't robot properly on an empty stomach, right?

5 PM arrived just on time, with our team having everything set up. Jacky of course delivering his talk noticeably well, it's now coming down to me, my one job...
To run Clara (darwinOp2) to perform weight-lifting in front of the live audience.

It ran pretty well the day before, it ran just moments before the presentation. But hey, it's not computer science unless it breaks down when the camera starts rolling...
Something froze up, the program wasn't starting. Luckily Jacky covered us with a brief segment on the importance of the scientific attitude...
Murphy's law at work!
Instead of having the attitude of trying to say [you] want to do the minimum amount of work to prove your system sortof works [. . .] you must try to make your system fail. If you try really hard to break your system, and you can, then you have something interesting. ~Jacky Baltes

Sitting in the comfort of the lab, or at home working alone, doesn't match the experience of problem-solving live in front of an audience. Despite the hiccup, we got things working -- with many thanks to Kiral for speeding up the process. Although the demonstration was not flawless by any means, I feel I've learned a lot from the experience and will focus on ensuring issues like this are minimized in the future.







Looking forward to the months to come; however I best wrap up this blog entry, as I'm sitting here alone in the lab writing away text on a day off...
A day off? What on earth are you doing here on a day off writing blog posts for?
     You're absolutely right, the robots need work!
         

No comments:

Post a Comment

Please feel free to give feedback/criticism, or other suggestions! (be gentle)

Subscribe to Updates via Email