This is great! I used to work at Savioke (Relay Robotics) where we collaborated with all major elevator companies to develop such APIs. Most of them, however, could not offer it to us even if they had an API, simply because the elevators were too old. By 2019, TK Elevators had by far the best API -- and I definitely biased because they pretty much implemented the spec we had given them. Still, the vast majority of elevators that we enabled our robots to ride, were integrated using a custom control board we had developed. And engineering, software or hardware, was not the issue. The main problem was that no elevator technician had ever done such an installation before and it took them months to get internal approval for it. So the real task was that of educating these companies and getting executive buy-in. That process took and presumably still takes time, meaning that robots riding elevators is not fundamentally a technical problem.
This is great! I used to work at Savioke (Relay Robotics) where we collaborated with all major elevator companies to develop such APIs. Most of them, however, could not offer it to us even if they had an API, simply because the elevators were too old. By 2019, TK Elevators had by far the best API -- and I definitely biased because they pretty much implemented the spec we had given them. Still, the vast majority of elevators that we enabled our robots to ride, were integrated using a custom control board we had developed. And engineering, software or hardware, was not the issue. The main problem was that no elevator technician had ever done such an installation before and it took them months to get internal approval for it. So the real task was that of educating these companies and getting executive buy-in. That process took and presumably still takes time, meaning that robots riding elevators is not fundamentally a technical problem.