Is development still continuing? 

so you just made your own gimbal controller?

why didnt you use a cots gimbal

That's basically accurate! :)

However, as I understand it (at least on the cheap ones I've seen), gimbal controllers don't have the magnetic encoder for closed loop position control and just open loop do sinusoidal drive of the BLDC. This is exactly what we do for the calibration step.

wait so hes even using the magnetic encoders xD

so he literally did nothing special

typical hackaday

maybe try reading the back emf voltages for position information, much faster and may not need prediction like shown here: http://electronoobs.com/eng_circuitos_tut4.php

Yes, that's the plan for future versions of the software. However back emf sensing doesn't work when the motor is not spinning.

Can this be upgraded to work with larger motors? cause I am looking for a controller for a electric motorcycle and would love it if it was a diy project..

The 5208 is pretty nice but for driving larger wheels and heavier loads would you consider what's needed to drive a 190Kv 6364 or 6374 ?
I'm hoping its just the Kv nad the mosfet size (675W vs 2450W, 22V max vs 37V max) ??
e.g. https://hobbyking.com/en_us/turnigy-aerodrive-sk3-6364-190kv-brushless-outrunner-motor.html

Maybe you would benefit from a planetary gear reductor where the motor shaft would be the inner gear?

Q&A

- Why did you (the team) need slow motion control?

We wanted to make really cool balancing robots. Cool meant high tech and quiet, which means use BLDC because that's the new fancy stuff. But BLDC controllers didn't work at low velocity. We'd been playing with many off the shelf controllers and knew there wasn't anything special. One killer trick combined with cheap drive electronics to make an awesome little package.

- Why is it important for the industry to have an open source version of this?

Because the entire industry is locked up with really expensive and closed controllers. There is little drive to innovate. We'd like to see that change. Showing the world it isn't "black magic" to make it accessible for all.

- How does this compare to VESC?

We're using fewer, more cost-effective components, and we have fewer features. We want to make it so everyone can make a bunch of boards for cheap.

- How does this compare to dynamixel in cost and performance?

Much higher power density. Faster speeds.

- What else will be added to the project, down the line?

Current sensor. Back-EMF sensing. Open control schemes.