If you have a robot that is controlled by 

- seeing light and dark on some area of a cell phone screen, and 

- that cell phone is using a remote meeting app (google meet, zoom, whatever...) and 

- the person on the other end has a card with a white area in the middle and black on both sides

then they can drive the remote cell phone around by moving the card left and right, up and down. 

That's too simple to work, right? Right?

What if you tool an old Roomba, and used the serial interface to connect a Pico, and then had a selfie stick mounted to the top with a cell phone that people on the remote end started a video call with... 

https://wokwi.com/projects/392667639628269569