Discovering Robotics

In his 1990 paper, "Elephants Don't Play Chess", Rodney Brooks, argued to be the father of modern robotics, states that in order for robots to accomplish tasks in an environment shared with humans, their higher cognitive abilities need to be based on the sensory-motor coupling. That is to say that intelligence cannot transcend from the existence of a body that senses from the external environment and acts on it through movement; an idea shared by the neuroscientist Antonio Damasio in his beautiful book "Descartes' Error: Emotion, Reason, and the Human Brain".

I have always been fascinated by the complex functioning of the human body and I believe that robotics, in its clumsy attempt to emulate it, has a fundamental role not only in increasing our quality of life but also in deepening the understanding of how we function.

I was 19 when I read "..my first idea was a robot arm, with a camera mounted in the ceiling above, that would interact with objects placed in front of the arm." (R. Brook, "Flesh and Machines: How Robots Will Change Us") and started playing at home with a 4DOF robotic arm (Lynxmotion) equipped with vision through an off-the-shelves camera. The robot used an  Arduino as a buffer to communicate with my PC and email me snapshots of my room when it detected motion (Video 1).

I later added a muscle sensors (MyoWare) and a 3-axis accelerometer to control the movements of the robotic arm (Video 2): tilting and panning my forearm would move the robotic arm in the same directions while contracting the extensor muscle on the forearm would cause the gripper to close, allowing to tele-manipulate the device and grasp objects. 

Interfacing the human nervous system with a device made out of aluminium and silicon was both compelling and intriguing; this curiosity has been my drive ever since.

Video 1: motion detection.  The robotic arm could detect motion and scan its surroundings, alerting me with an email. Its eyes, crucial for vision, were taken from Toy Story's epic character Potato Head.  

Video 2: EMG-driven grasping.  A 3-axis accelerometer and a muscle sensor on my forearm allowed me to control the position of the gripper and to grasp objects. The video ends with an high five.