Sometimes A Bot Just Has To Break The Rules

Robot Carl lives by very strict rules - No speeding and no side effects (one program affecting another).

After robot Dave got his SNES gamepad working, he wanted to go for a long walk and planned to walk as far and as fast as he was able. His new gamepad had two speeds: 0.1m/s (wheels turning at 180 degrees per second) or 0.175 m/s (wheels turning around 315 DPS). Dave can go as fast as 0.24m/s but tends to veer off straight for some reason.

To get ready for his walk, he appropriated a copy of robot Carl’s “wheel log” software to keep track of how far he was walking and how much his head was spinning from the exercise.

A few diversionary learning experiences later, Dave remembered his walking goal but for some reason was always feeling tired, unable to go any faster when the SNES turbo button was asking him to run as fast as he could.

Today, Dave’s mystery tiredness was shown to come from the fact robot Carl always obeys a 150 DPS speed limit, and the wheel log program Dave lifted from Carl, had a side effect of not letting Dave exceed that speed limit. Thus normal and turbo were both only 150 DPS, and not 180 and 315 as commanded.

One small matter of program change to the wheel logger and Dave is now dreaming of being the first GoPiGo3 robot to “run a 1k time trial” (estimated to take 1:35).


Dave’s first “training run” was this morning, but I could only entice him to run 2 meters!

He wanted to check the accuracy of his “sport watch”, so he raced across the office traveling 2.083 meters (the best I could do with a 2 meter tape measure …), then asked “What does my wheel log show?”

wheel.log entry:

2021-09-27 09:21|[]travel:  2089.8 rotation:     0.3  motion:    15.2 sec  heading_est:     0.3

Considering Dave’s dusty tires are not very sticky material, “running” on a glossy slick wood floor, to claim 2.0898 meters is only 0.3% error. That is way better than I expected.

With this training run under his belt, Dave is now asking for “running shorts” so he can run on the sidewalk outside without finishing with a case of “robot jock itch” (dust and metal particles on his encoder magnets).

Dave was pleased with his “performance” until I told him his 1k pace of 2:05 was 32% slower than the “best in GoPiGo3 class” seen on the “robot running forum” discussions. (He rejoined that his coach told him to walk a little to warm up before running at full speed. )

Dave Asking To Go Out

1 Like

Wow - that’s amazing.

Good luck to Dave on his upcoming runs.