thinking in rules

A Vision for Home Robotics

Loren Goodman | 12/5/2013



Ever since I was a kid, I have had this vision for how robots would find their way into our homes. While I was growing up, the robotic road map from pop-culture and the media was always focusing on the progression of robots towards a humanoid form. Even today, if you search for “robotics” in google images, everything that comes back looks like a human, an assembly line, or a pet straight out of some Tim Burton movie. The humanoid style robots never really made sense to me; I just don’t think robots are going to go that direction; my vision is an evolution from the opposite direction. The first airplanes were designed around making wings that could flap, which, according to middle-of-the-night-history-channel, turned out to be a flop. The Wright brothers realized that flying was the goal, not looking like a bird. The Roomba does not look like a robot or a vacuum cleaner: its form arose out of its function.

The value proposition of robots will not be defined by how well they resemble humans, it will be defined by how well they perform tasks on behalf of humans, how easy they are to maintain, how easy they are to instruct, and how smoothly they integrate into other human owned technology. The lines are already starting to blur between what is a robot and what is an appliance. Our dishwashers, vacuums, and toasters are all designed to replace human labor. Are these robots?

Rubik Cube RobotPicture a Rubik’s cube with all the pieces taken off except the center squares on each side. And suppose that each of the six core posts have a piston in them which allows the post to extend and retract. At the end of each arm, instead of the colored sticker, there is an electro-magnet. Both the pistons and the magnets are computer controlled. By managing the polarities of the magnets, the cubes can be forced to stick to one another and they can also be forced to rotate around each other in a stepwise fashion akin to how electric motors work today. Each cube is capable of identifying itself to the control system with what cubes are connected to it and what direction it is facing.

A computer has a three dimensional understanding of the entire structure based on the cubes and what they are connected to. A pile of these cubes would be capable of assembling themselves into a variety of shapes by simply rotating around each other until the desired position is achieved and locked in. For example, if I directed the computer to build a large tower, cubes would keep climbing up the outside, like ants going up a wall, until they reach the top. Even though the cubes are moving, there is a physical connection among the cubes all the way back to the control system which provides communication and power.

Imagine if you had a few thousand of these cubes in your home. You could tell them to shape themselves like a couch, then sit down on it, watch a movie, and then tell the couch to put itself away. You could have them become the kids table for Thanksgiving dinner and then turn into a fort for the kids to play in after dinner. By synchronously extending the pistons, a set of cubes can lift themselves up, making room for additional cubes to roll in underneath and lift them even further. If you want to reach something on a high shelf, you can stand on them and have them lift you up. If you want to move some heavy furniture, they can climb underneath and start jacking it up.

Where I think it gets really interesting is the introduction of specialization. With just pistons and magnets, you can create furniture and get a helping hand. Suppose there were cubes with specific features. For example, if one side of a cube was a display, they could arrange themselves in a grid and be a television. Cubes with cameras and range finders would let you take 3D scans and pictures. Cubes with heaters and temperature sensors could replace your furnace by simply going into a cold room and heating it up. The cubes that were the toaster in the morning could be reformed as the oven for dinner. Now add cameras, speakers, motors, tires, etc. Ultimately, any electronic device could be assembled from a set of cubes with specialized features and on a whim reassembled into almost anything else at minimal cost.

TransistorRobotic cubes are the next transistor. A single transistor on its own might appear to be nothing more than a simple switch, but when they are connected and coordinated these simple transistors have enabled the creation of every piece of digital technology we own. A lot of simplicity works better than a lot of complexity. The first transistor was made in 1947 and it was ½ inch on each side. Today you can put 100 million of them on the head of a pin. If robotic cubes follow in those footsteps, they might be the size of golf balls today, but tomorrow, the size of marbles, and in fifty years, the size of dust.

Today you have things like gas, water, cable, and electric coming into your house. Why not a self-assembling robotic dust feed that lets you use and return specialized cubes as needed? This would be the ultimate in recycling! With a connected supply chain coming into your home, turning your couch into a vibrating massage table is as easy as downloading an app on your phone today. Every month you get a bill for what you used and every month the manufacturer gets paid for what they sold.

Need a robotic exoskeleton for a broken leg, no problem. Want to have surgery at home via remote control? No problem. Want to convert your living room into the holodeck? Engage!

There is a lot of tech and self-assembly research out there that is capable of making this really work. Today, people make life-size cars out of Legos for fun; tomorrow, we will drive them. I can hardly wait!

comments powered by Disqus