A few months ago I had reason to reverse-engineer the Canon EF lens communication protocol. Eventually I was able to remotely control the lens aperture, which was my goal. I know there's a low-level interest in this sort of thing, and I'm hoping I can make contact with people who are able to fill in some of the gaps.
Obviously Sigma, Tamron, Tokina etc. have reverse-engineered the protocol, but also niche companies such as Redrock Micro and Birger. Since the protocol is over 20 years old I was surprised to find little detailed information online.
I've attached an example of my experimental set-up, one that accesses the lens pins via a modified extension tube. The microcontroller based circuit controls the lens by passing serial command/requests - essentially replacing the camera body. The data displayed on the LCD came from the lens.
For those interested, here's Canon's official statement regarding third-party lenses. It provides some insight as to why Sigma and friends sometimes get it wrong,
I've written up my findings in a pdf. As a courtesy to Canon I asked them to what extent I can publish this 'proprietary intellectual information' online. Sadly, after 1 month, their `Legal Team' has not returned the courtesy. I won't be posting full details until I hear their opinion. I am, however, keen to hear your ideas and discuss applications.