A group of researchers at UC Berkeley (including EE Prof. Sergey Levine, grad student Tuomas Haarnoja and undergraduate researcher Aurick Zhou) and Google Brain have used maximum-entropy reinforcement learning to make a quadrupedal robot teach itself to walk. It taught iself through trial and error in a mere two hours before researchers introduced the machine to new environments, like inclines and obstacles, where it adapted with ease.
CS Prof. Ken Goldberg and members of the AUTOLAB including postdoc Jeffrey Mahler (Ph.D. '18), grad students Matthew Matl and Michael Danielczuk, and undergraduate researcher Vishal Satish, have published a paper in Science Robotics which presents new algorithms to compute robust robot pick points, enabling robot grasping of a diverse range of products without training. They trained reward functions for a parallel-jaw gripper and a suction cup gripper on a two-armed robot, and found that their system cleared bins with up to 25 previously unseen objects at a rate of over 300 picks per hour with 95 percent reliability.
EE Assistant Prof. Rikky Muller has been named an IEEE Solid-State Circuits Society (SSCS) Distinguished Lecturer (DL) for the two year term of 2019-2020. SSCS DLs are experts in current integrated circuit technologies who are chosen to speak at chapter meetings and regional seminars because of their skills as deeply knowledgeable and excellent communicators. Muller co-directs the Berkeley Wireless Research Center (BWRC) with Prof. Elad Alon, who is currently serving the second year of his term as a 2018-2019 SSCS DL. Muller will be a featured speaker at the Berkeley EECS Annual Research Symposium (BEARS) on February 14, 2019
For the first time ever, an international team of researchers--led by Prof. Sayeef Salahuddin--imaged the microscopic state of negative capacitance. This novel result provides researchers with fundamental, atomistic insight into the physics of negative capacitance, which could have far-reaching consequences for energy-efficient electronics. “The upshot is that the opposite relation between charge and voltage could locally enhance the voltage across the common dielectric material,” said Salahuddin. “The voltage ‘amplification’ gained could be used to reduce the supply voltage requirement in a transistor, thus making computers and other electronic devices more energy-efficient.”
CS alumni Pulkit Agrawal (M.S. '14/Ph.D. '18, adviser: Jitendra Malik) and Jacob Andreas (Ph.D. '18, adviser: Daniel Klein ) will join the MIT Department of Electrical Engineering and Computer Science, and EE alumna Cathy Wu (Ph.D. '18, adviser: Alexandre Bayen) will join the MIT Department of Civil and Environmental Engineering, as assistant professors. Agrawal, who co-founded SafelyYou, Inc., studies topics spanning robotics, deep learning, computer vision, and computational neuroscience. Andreas, who was a member of the Berkeley Natural Language Processing Group and the Berkeley Artificial Intelligence Research Lab, focuses on using language as a scaffold for more efficient learning and as a probe for understanding model behavior. Wu, who was recently awarded the Milton Pikarsky Memorial Award, focuses on research involving machine learning, robotics, intelligent systems, and mixed-autonomy mobility.
Research collaboration between Singapore–Berkeley Building Efficiency and Sustainability in the Tropics (SinBerBEST 2 -- part of the Berkeley Education Alliance for Research in Singapore center) led by EE Prof. Costas Spanos, and UC Berkeley’s Center for the Built Environment (CBE) resulted in three 2018 Building and Environment Best Paper Awards. The awards, which are presented annually by the Building and Environment journal, recognize originality, contributions to the field, quality of presentation, and soundness of science. The honored papers were titled "Automated Mobile Sensing: Towards High-Granularity Agile Indoor Environmental Quality (IEQ) Monitoring," "Development of the ASHRAE Global Thermal Comfort Database II," and "Personal Comfort Models: Predicting Individuals’ Thermal Preference Using Occupant Heating and Cooling Behavior and Machine Learning."
Recent EE alumna Cathy Wu (Ph.D. '18, adviser: Alexandre Bayen) has won the Council of University Transportation Centers (CUTC)’s Milton Pikarsky Memorial Award. This honor, which recognizes the best Doctoral dissertation in the field of science and technology in transportation studies each year, was awarded for Wu's dissertation titled “Learning and Optimization for Mixed Anatomy Systems – A Mobility Context.” “This is a great honor,” says Wu. “I feel very fortunate to have been able to study optimization challenges at very different layers of our complex transportation systems in the context of self-driving vehicles, from congestion to routing to questions touching on planning and policy.”
CS alumnus Ken Thompson (B.S.‘65/M.S.‘66) is a member of the 2019 class of inductees into the National Inventors Hall of Fame for creating the UNIX Operating System. Thompson and Dennis Ritchie's creation of the UNIX operating system and the C programming language were pivotal developments in the progress of computer science. Today, 50 years after its beginnings, UNIX and UNIX-like systems continue to run machinery from supercomputers to smartphones. The UNIX operating system remains the basis of much of the world's computing infrastructure, and C language -- written to simplify the development of UNIX -- is one of the most widely used languages today.
A new neurostimulator, described in a paper co-authored by EE Prof. Jan Rabaey, Prof. Jose Carmena, Assistant Prof. Rikky Muller, grad students Andy Zhou, George Alexandrov and Ali Moin, and alumnus Fred Burghardt (B.S. '92/M.S. '94), in the journal Nature Biomedical Engineering, works like a “pacemaker for the brain" to both monitor electrical activity and therapeutically stimulate electric current to the brain at the same time. The device, named the WAND, could potentially deliver fine-tuned treatments to patients with diseases like epilepsy and Parkinson’s. Muller's research is part of the CZ Biohub, a division of the $5 billion Chan Zuckerberg Initiative. Rikky Muller and Jose Carmena are both scheduled to present their work at the 2019 BEARS symposium in February titled "The Future of Medicine: An EECS Perspective."
All four students who were nominated for Computing Research Association (CRA) Outstanding Undergraduate Researcher Awards in 2019 were recognized: Dibya Ghosh (nominated by Sergey Levine), Hong Jun Jeon (nominated by Anca Dragan), and Jonathan Lee (nominated by Ken Goldberg) were named as finalists, and Annie Xie (nominated by Sergey Levine) was named a runner up. The CRA award program recognizes undergraduate students in North American colleges and universities who show outstanding research potential in an area of computing research.
