Applying Biological Principles in Designing Robots of the Future

Hod Lipson is noted for bringing biologically-inspired approaches to the study of robotics - all in an effort to answer what he believes to be two great challenging questions in engineering: "First, can we design machines that can design other machines, and second, can we make machines that can make other machines?" he asks.
Both of these questions, Hod contends, lie at the crux of understanding the engineering process itself, and science's ability to design, make and maintain complex machines, like robots, in the future.

Biological life itself, he says, has answered many of these questions better than the best teams of human engineers ever could -"which is why I use biologically-inspired approaches in my work, as they bring new ideas to engineering and new engineering insight into biology."

Hod is director of Cornell University's Computational Synthesis Lab (CCSL) at the Sibley School of Mechanical and Aerospace Engineering, Ithaca, N.Y. He has led work in areas such as evolutionary robotics, multi-material functional rapid prototyping, machine self-replication and programmable self-assembly.

"In robotics, I am specifically interested in the questions of how we can make machines more adaptive to the environment, to other machines and to changes in themselves, such as system failures," Hod explains. Robotic systems today are superhuman in their accuracy, in their speed, in their ability to work 24/7 in hazardous environments, he says." But, their inability to adapt to new situations is really their weak point. In contrast, biology is very good at adaptation. As robotic environments and tasks become more complex, it will eventually boil down to adaptation, which is highly important to the sustained operation and long-term viability in robots."

In exploring these biologically-inspired questions in robotics, what has this taught Hod about humans and our own evolution?

"When you study robotics, it forces you to rethink, in a very quantitative way, the attributes we hold close and consider unique in our definition of what it means to be human," he says. "For example, what is creativity? If machines can create new things and ideas that infringe on patents (which humans have traditionally defined as being creative), what does that mean about creativity?" Similarly, he says, when we have computers that can generate experiments and ask questions, what does that mean about curiosity?

"When you actually work with robots trying to emulate these very characteristics, it forces you to think about these traits in a very precise and quantitative way. Ultimately, I think it leads to deeper questions and better understanding of these concepts."

What types of tasks do you think a 'self aware' robots could do?

Hod Lipson: "Adaptive and Self-Reflective Systems" at Foresight 2010 Conference from Foresight Institute on Vimeo.

Read more about Hod Lipson here