ZurichRobotics #4
Seif Ismail from Watney Robotics on Conway's Law in robotics and Victor Klemm from ETH Zurich on designing robots with legs, wheels, and arms.
Loading…
181 registered
Speakers
Seif Ismail
Founding EngineerWatney RoboticsVictor Klemm
PhDETH ZurichAbout
Seif Ismail (Founding Engineer, Watney Robotics): Conway's Law in Robotics Over the past two years building and deploying robots at Watney, we arrived at a set of counterintuitive lessons that fundamentally reshaped how we think about robot systems. Deployability and scalability became the driving principles behind every technical and organizational decision: hardware architectures optimized for serviceability, software systems prioritized simplicity, robustness, and generalizability over novelty and operations evolved into a first-class engineering discipline rather than an afterthought. Deploying useful robots requires complete alignment at all company levels, and the real-world constraints that come with it paradoxically produce better research and engineering outcomes. Drawing from concrete deployment experiences, this talk argues that the future of scalable robotics will not be determined solely by breakthroughs in model capability or hardware sophistication, but by organizations capable of aligning every function—from engineering to operations—around the realities of deployment.
Victor Klemm (PhD, ETH Zurich): Robots with Legs, Wheels, and Arms: Controlling and Designing Robots at RSL that move purposefully
Legged, wheeled, and manipulating robots are increasingly capable of operating in complex, real-world environments. This talk presents research from the Robotic Systems Lab (RSL) at ETH Zurich on the development of robots that combine locomotion and manipulation. Over the past years, the field has undergone a profound paradigm shift: learning-based methods, trained at massive scale in simulation, have replaced many classical control pipelines and unlocked levels of robustness and agility that were previously out of reach. Simulation and massive parallelization have become central in this process, rendering challenges in bridging the gap between simulation and reality increasingly important.