Trendy sensors, microcontrollers, and PID (proportional-integral-derivative) management have made self-balancing robots and autos fairly accessible. That’s very true when steadiness is barely required in a single axis, like with a hoverboard. But it surely does get harder when you should transfer and steadiness in two axes. Just a few years in the past, James Bruton tried to drag that off with a robotic that had a single omni wheel, however that didn’t work very properly. By no means one to surrender, he revisited the thought to construct a brand new and utterly reengineered model of the only omni wheel self-balancing robotic.
Conventional omni wheels are solely pushed within the “regular” axis. A motor can rotate them ahead or backward like a standard wheel, however the small rollers that permit for side-to-side motion aren’t pushed. That’s purposeful, as a result of they permit fascinating motion when 4 omni wheels work collectively in live performance. Nevertheless, in a case like this the place the robotic solely has one wheel, free-spinning rollers could be counterproductive.
For that motive, Bruton designed and constructed a particular type of omni wheel that has pushed rollers. One motor rotates your complete omni wheel within the “regular” axis, then a second motor rotates the rollers perpendicular to that through a collection of belts and pulleys in an association considerably much like a planetary gearbox in look. As is often the case, Bruton 3D-printed a lot of the customized mechanical components of the robotic. Although this additionally has CNC-cut plywood sandwiched collectively to assemble the massive “hub” gear.
A Teensy 4.0 microcontroller improvement board controls the robotic, utilizing an IMU (inertial measurement unit) to watch motion of the robotic. It controls the ODrive motors by means of a CAN (controller space community) bus. Energy comes from a set of enormous hobby-style lithium battery packs.
With a self-balancing robotic like this, success comes all the way down to correct PID management. Bruton took benefit of an Arduino PID library after which put loads of effort into PID tuning. Your complete level of PID management is to enact compensation for tipping, with out overcompensating and creating an more and more extreme oscillation. On this case, Bruton had to try this in two axes working collectively.
The outcomes are fairly promising. It doesn’t actually drive round — Bruton did attempt to drive it round along with his common distant, however that wasn’t dependable. Nevertheless, it does stand nonetheless on one spot very properly whereas balancing. If Bruton provides it a shove, it’s going to appropriate itself to remain upright.
