ABSTRACT
Computational techniques involving modeling, simulation, and numerical optimization have supported research and development for years. In this talk, I will show how some of these techniques can be leveraged for simplifying complex design problems in the field of robotics.
In particular, I will present design tools that take advantage of these techniques, and enable novices to design physical structure and functional behavior of diverse robotic devices. With the help of parameterized domain-specific models and physics-based simulation that can encode complex design aspects, these tools democratize the robot design process. These tools also incorporate efficient algorithms that can search high-dimensional design spaces at interactive rates for user-preferred solutions, and provide easy-to-use visual interfaces that can support user-in-the-loop interactive design. As a validation of capabilities of these tools, I will show various fabricated robot prototypes and results of user-studies conducted with novices. By supporting rapid creation of custom robots for the people and by the people, these tools help take a step towards the future of ubiquitous robotics, wherein robots would be integrated in daily human life just as smart phones and computers are today.