CS Colloquium - Josh Bongard

Event time: 
Thursday, February 21, 2019 - 4:00pm
Location: 
AKW 200 See map
51 Prospect Street
New Haven, CT 06511
Event description: 

CS Colloquium - Josh Bongard

Refreshments available at 3:45

Host: Marynel Vazquez

Title: Sim2real, sim2life, and sim2all: designing and manufacturing non/living systems.

Abstract:

Advances in physical simulation, optimization, and manufacturing processes, when combined, can yield pipelines for automatically designing and building machines and living systems. I will present three of several such pipelines we have built to date. The first combines soft robot simulation with physical robotic skins to yield shape-shifting machines. The second combines an aerodynamic simulator with a flight test rig to yield ornithopter wings. The third pipeline dictates ways to combine living tissues to yield novel living systems that lie somewhere between artificial machines and organisms. I will also discuss our crowdsourcing work, in which non-experts can collectively guide pipeline flow toward safe robot designs and behaviors. I will conclude with our plans to unite these pipelines into a single tree: designs become increasingly embodied and differentiated as they flow up the branches to new manufacturing processes continuously added as leaves.

Bio:

Josh Bongard is the Veinott Professor of Computer Science at the University of Vermont and the director of the Morphology, Evolution & Cognition Laboratory. His work involves computational approaches to the automated design and manufacture of soft-, evolved-, and crowdsourced robots, as well as living systems. A PECASE, TR35, and Microsoft New Faculty Fellow award recipient, he has received funding from NSF, NASA, DARPA, the U.S. Army Research Office and the Sloan Foundation. He has authored 32 and 74 peer-reviewed journal and conference publications respectively, and is the author of the book How The Body Shapes the Way we Think. He runs an evolutionary robotics MOOC through reddit.com and a robotics outreach program through twitch.tv.