Advancements in the field of robotics are fueled by research, which in turn heavily relies on effective platforms to test algorithms for robot control and navigation. While numerous robotics platforms have been developed over the past decades, most of them have shortcomings that limit their use in research settings.

Researchers at the University of California (UC) Berkeley recently developed Berkeley Humanoid, a new robotic platform that could be used to train and test algorithms for the control of humanoid robots. This new humanoid robot, introduced in a paper posted to the preprint server arXiv, addresses and overcomes some of the limitations of previously introduced robotics research platforms.

“Having conducted several experiments with commercially available robots, we have become aware of some of their weaknesses,” Qiayuan Liao, co-author of the paper, told Tech Xplore. “For instance, some robot hardware is very expensive, while other hardware is not designed especially for learning-based control or for research, which often means that it is ‘fragile,’ easy to break, and hard to maintain and repair.”

Recently, computation in memory becomes very hot due to the urgent needs of high computing efficiency in artificial intelligence applications. In contrast to von-neumann architecture, computation in memory technology avoids the data movement between CPU/GPU and memory which could greatly reduce the power consumption. Memristor is one ideal device which could not only store information with multi-bits, but also conduct computing using ohm’s law. To make the best use of the memristor in neuromorphic systems, a memristor-friendly architecture and the software-hardware collaborative design methods are essential, and the key problem is how to utilize the memristor’s analog behavior. We have designed a generic memristor crossbar based architecture for convolutional neural networks and perceptrons, which take full consideration of the analog characteristics of memristors. Furthermore, we have proposed an online learning algorithm for memristor based neuromorphic systems which overcomes the varation of memristor cells and endue the system the ability of reinforcement learning based on memristor’s analog behavior.

Full abstract and speaker details can be found here: https://nus.edu/3cSFD3e.

Register for free for all our upcoming webinars at https://nus.edu/3bcN9pS.

Algorithms like OpenAI’s GPT-4 are like brains frozen in time. A new study shows how future AIs could learn continuously in response to a changing world.

The CEO of Amazon Web Services suggested in leaked audio that human devs may not be coding in as little as two years.