Embedded Processing

I have an intense interest in autonomous systems, especially ones that incorporate learning to refine how they sense and interact with the world. Before becoming a technical editor at EDN, I devoted many years to research and development designing and building fully autonomous vehicles. A certain mission, which happened a few weeks ago, involving a satellite in orbit, had a personal significance to me, as I was part of the team that spent a few years working on the original technology that powered the final autonomous portion of that system. With systems as complicated as these, it can be many years, potentially decades, between the original work and a working demonstration in field conditions.

Boston Dynamics released a new demonstration video of its larger quadruped robot operating in field conditions. I am very excited by what I see in the video as it demonstrates how far the technology has progressed over the past few decades of research. In addition to the lower quality but faster-loading demonstration video, the company's website offers a high-quality video. There are other high-quality videos of other autonomous robots at the website, including similar robots with four and six legs.

Of particular interest in the BigDog video is that much of it takes place in field conditions with slopes and real-world ground conditions, such as going up and down hills covered with leaves or even snow. There are a few slow-motion replays in the video that highlight particularly interesting behavior by the system. The first slow-motion sequence involves a person giving the robot a strong lateral shove with their foot; we see how the robot recovers from it. The next slow-motion sequence is even more impressive, as the robot walks onto an ice patch and recovers; of special note to me was how the robot used its "knees" to recover from sliding on the ice. The video finishes up with some lab demonstrations of this 340 pound (when fully loaded) robot traversing a cider block mound, a rock bed, and leaping over a section of the floor.

Behaviors like these are too complicated to program explicitly; they rely on the system learning how to interact with the world and compensate for the variability in the world. However, I do not believe this robot is fully autonomous in that I think there was an operator directing the robot at the level of turn right or left and go forward via a wireless controller. This in no way diminishes the accomplishment of this team, as the more interesting autonomous portion of the system, at this point of time, is the one adjusting the legs to compensate for unexpected forces from the environment, such as a slippery surface, uneven surfaces, moving surfaces, or even strong lateral forces. One thing I would have liked to see in the video was the robot recovering from a non-injurious fall.

The potential uses for this type of technology are vast. An obvious use scenario is to transport food, clothing, supplies, and medicine to people in dangerous war-torn areas while risking fewer people's lives. This type of technology can be applied to smaller robots that can more quickly and safely search for survivors in rubble, from say a fallen building. Additionally, this type of autonomous system, or subsystem, is analogous to the work being done in the automotive industry to make smarter vehicles.

It is my plan, if there is enough interest, to pursue with Boston Dynamics, and any other company working on these types of projects that might be willing to share some information about their systems, a set of prying eyes features that will highlight the various subsystems that are necessary to make these types of systems operate successfully. I am particularly interested in sharing with you what it takes to seed and maintain the learning algorithms that I expect this type of system uses. Take this opportunity to post what you would like to see in these Prying Eyes features.