Webinar

In this webinar, Matt Brown, Stanford Dynamic Design Lab, discusses techniques and design considerations that enable a controller to simultaneously handle lateral stabilization and collision avoidance.  Fully handling emergency collision avoidance requires an ability to coordinate both steering and brakes/throttle; directly modeling the coupling between lateral and longitudinal tire forces enables the controller to handle very difficult scenarios that require both limit steering and limit braking.

John Alsterda, Ph.D. candidate in the Dynamic Desing Lab will present Contingency Model Predictive Control (CMPC), a motion planning and control framework that optimizes performance objectives while simultaneously maintaining a contingency plan – an alternate trajectory that avoids a potential hazard. By preserving the existence of a feasible avoidance trajectory, CMPC anticipates emergency and keeps the controlled system in a safe state that is selectively robust to the identified hazard.

In this webinar, computational techniques for enabling self-driving cars to make predictions of the social world around them will be discussed, focusing on the rapidly-evolving field of trajectory forecasting. The webinar will feature three segments, beginning with a discussion about the problem of trajectory forecasting and methods for solving it. The second part of the webinar will discuss how, even with the ability to predict the future, it is still unclear how best to integrate such trajectory forecasting models within the autonomy stack.

Autonomous vehicles such as self-driving cars and Unmanned Aerial Vehicles (UAVs) require vehicles to navigate at low altitude and in urban environments posing positioning challenges, such as blockage and multipath effects for GPS signals and unreliable features for vision or LiDAR-based navigation. In this talk, Grace Gao, assistant professor, aeronautics and astronautics will present an overview of her lab's recent work on reliable and safe positioning and navigation for autonomous systems.

John Subosits, Stanford Dynamic Design Lab, explores what effects must be captured for a vehicle to consistently drive at the limits of friction. Insights from this work can be used to design algorithms that operate over the full range of vehicle performance, maximizing an autonomous vehicle's ability to operate skillfully when racing or safely when confronted with an emergency.

Karen Leung, Stanford Autonomous Systems Lab, will first introduce backward reachability analysis and then show how it can be used to construct a minimally-interventional safety controller operating within an autonomous vehicle control stack with the role of ensuring collision-free interaction with an externally controlled (e.g., human-driven) counterpart while respecting static obstacles such as a road boundary wall.

Autonomous Mobility-on-Demand (AMoD) systems, wherein a fleet of self-driving vehicles serve on-demand travel requests, present a major paradigm shift in modern mobility and a unique opportunity to alleviate many of our transportation woes. In this talk, Ramon Iglesias, Ph.D., Autonomous Systems Lab, presents a queuing network approach to the problem of routing, rebalacing, and charging a fleet of self-driving vehicles within an AMoD system.

What does a barrel falling out of a warehouse in the 1800s have to do with liability for emerging automated vehicles (AVs)? Bryan Casey - Lecturer at Stanford Law School and CARS Legal Fellow - will argue that a string of cases dating back to the industrial revolution actually provides an elegant solution for resolving complex questions of fault that lie at the heart of AV accidents.

What do we do with the data we get from the multiple sensors companies create? Interventions do not evolve organically from sensors. Creative interventions can inform the future research on sensing. Cases and ideas from Pablo Paredes' lab: steering wheel that senses stress and a chair that guides users towards breathing interventions.

The goal of Dorsa Sadigh, assistant professor in computer science and electrical engineering, research is to enable such human-robot systems to be safe and interactive. Sadigh aims to develop a formalism for design of algorithms and mathematical models that facilitate correct-by-construction control for safe and interactive autonomy.