Sio

Robots have the potential to solve big problems. They already assemble cars and explore planets far away, but they are usually kept away from humans. What stops a robot from helping you clean your dishes or carry the moving boxes? It is the fact that they do not fully perceive their environment. Usually dependent on a single vision system, they do not know where you are, which makes their movements potentially harmful and also means they cannot respond to human input such as touch.

SIO is a skin that gives robots both a sense of vision and a sense of touch. This enables them to perceive and work alongside humans. If it senses that a human is in its way, it knows to stops. If a box is slipping out of its gripper, it knows to use more force.

SIO works analogous to the compound eyes of a bee, meaning that it does not sense only through a pair of eyes, but through many more. Because the system can be integrated with the robot’s motion planning system, SIO eliminates the need for particular skills like programming to operate the robot. Some of the most interesting user interactions that become possible are shown in the video above.

In this group project, I was doing research and concept development, as well as proving that the proposed subsystems would work (proof of concept). While the video shown above has not got every subsystem working in an integrated fashion, we developed them as stand-alone prototypes. Particularly, we created a proximity sensing array and a custom force sensor, as shown below.