Orientation of Robots and Humans

                  Thanks to science fiction, any discussion about robots is bound to include a reference to machines that subjugate mankind due to their superior intelligence and strength. But the truth is that machines have a long way to go before they start bossing us around.

 Robots are great at certain tasks — the more uniform and repetitive the task, the more suitable it is for a robot. It’s when you switch things up that it gets tricky.

               Here’s an example. Let’s say you’re on a trip to a big city. You’ve got the skyline in view as you’re driving down the highway. Once you get off the highway and make a few turns, the same buildings you were looking at before are in a different orientation relative to your position. They look different but you recognize that they’re the same buildings and you can still use them as a reference point as you move around.

                   Robots have a rougher time of it. They cannot intuitively detect that they are looking at the same skyline — a programmer must build in that capability. There are lots of ways to do this: a GPS receiver, digital compass and virtual maps could help. But in an indoor environment those tools wouldn’t be as useful. Enter the computer scientists.

              A team of researchers at MIT lead by Julian Straub recently published a paper about an algorithm that can help machines detect and navigate new environments. It involves mapping out an area and framing it within the context of a virtual sphere. As the robot moves, the virtual sphere in the robot’s “brain” rotates to give the robot an updated frame of reference. By looking around, the robot should be able to map its location against the virtual representation.

              Imagine that every place you’ve ever visited, indoor or out, remained virtually mapped inside your brain, complete with every possible angle you’d perceive as you walk about or turn around within that environment. That’s what this algorithm simulates. It’s not necessarily a high-resolution approach — the maps are made up of frames of axes — but it gets the job done.

            This doesn’t mean that robots are going to become the world’s best tour guides within the span of a year, but it’s a promising development that could lead to amazing applications in the future. It might just take one more item off the list of things we humans can do better than robots. I guess I’d better prepare for our machine overlords.