Student projects done in the Jacobs Institute in a recent course on "Electric Mobility Engineering.
The above images are from co-PI and UCB alumnus Prof. Ankur Mehta at UCLA, demonstrating the ability to rapidly create low-cost prototypes of robotic arms or vehicles.
A UC Berkeley research team led by Prof. Sanjit Seshia has been awarded a four-year, $8.4M project by the Defense Advanced Research Projects Agency (DARPA) to research artificial intelligence-based approaches that augment humans to perform correct-by-construction design of cyber-physical systems (CPS). Cyber-physical systems tightly integrate computation with physical processes and have a broad range of societal applications, including in agriculture, aeronautics, civil infrastructure, energy, environmental quality, healthcare and personalized medicine, manufacturing, and transportation.
The award, made under a new DARPA program titledSymbiotic Design of Cyber-Physical Systems, aims to reduce the time from CPS inception to deployment from years to months and enhance innovation in design.
The Berkeley project, LOGiCS, (for Learning-Based Oracle-Guided Compositional Symbiotic Design of CPS) proposes a novel approach, blending AI and machine learning with guidance from human and computational oracles, to perform modular design of CPS such as autonomous vehicles that operate on the ground, in the air, and in water to safely achieve complex missions. Among the systems the Berkeley team will be working on are urban air taxis, autonomous underwater exploration, and autonomous multi-terrain search and rescue.
In addition to Seshia, the team includes Prabal Dutta, Björn Hartmann, Alberto Sangiovanni-Vincentelli, Shankar Sastry, and Claire Tomlin, all faculty in the Department of Electrical Engineering and Computer Sciences (EECS). Other members are two Berkeley alumni, Ankur Mehta at UCLA, and Daniel Fremont at UC Santa Cruz.
As DARPA specifies, the vision of the program is to vastly expand coverage and accelerate exploration of CPS design spaces with the symbiosis of two very different kinds of agents: humans with their uncanny ability to create intuitive associations across design domains, and machines with their ability to recognize statistical patterns from data and navigate vast search spaces for optimal solutions.
According to DARPA, "These systems and platforms integrate cyber and physical subsystems, and the enormous complexity of the resulting CPS has made their engineering design a daunting challenge."
Creating and manufacturing such systems requires hundreds of domain-specific tools orchestrated by large teams of engineers with extensive domain knowledge and subject matter expertise. To reduce the length of time needed to produce CPS and enhance innovative design, the DARPA program focuses on improvements in three areas:
-
Predictability: Improve the accuracy of predictions of performance before implementing software and physical components. This requires a balance between high-fidelity models that are cost- and time-prohibitive to produce and cost-effective modeling approaches that result in substantial uncertainty.
-
Convergence: Overcome discipline boundaries between design teams to address the interdependence of design decisions and accelerate convergence of a viable integrated solution.
-
Exploration: Encourage engineers to go beyond using only familiar and known-feasible approaches and explore design spaces that may contain unconventional but highly useful solutions.
The UC Berkeley EECS Department is part of the university's Division of Computing, Data Science, and Society, which leverages Berkeley’s preeminence in research and excellence across disciplines to propel data science discovery, education, and impact. It’s designed to meet the opportunities and demands of a world increasingly informed and shaped by data, machine learning, and artificial intelligence in virtually every arena, from health to business to politics; from our cities to our climate to the cosmos.