The robot revolution is a lot more advanced than most people realize. While robots are (correctly) assumed to carry out the will of their masters via detailed programs inserted into their memories, at least one Israeli researcher has developed a method for robots to “think” on their own. Professor Gal Kaminka, of Bar-Ilan University’s Department of Computer Science and Gonda (Goldschmied) Multidisciplinary Brain Research Center is one of Israel’s, and the world’s, leading contributors to intelligent robotics – the science of using artificial intelligence to make robots “smarter” – and this week he was awarded the prestigious Landau Prize for Arts and Sciences in the robotics category for his outstanding contributions to the advancement of science.
Landau Awards, given out annually by Mifal HaPayis (the people behind the Israel Lottery, which raised funds for educational projects) to five Israeli scientists who have achieved a recent breakthrough in their field of research, have gained international recognition, and contributed significantly to the advancement of science in Israel and abroad. Awards are given in each of five areas of research: natural sciences, exact sciences, social sciences, humanities and life sciences. The prize will be awarded at a festive ceremony on January 1, 2014.
Kaminka has a long history in robot research, especially in robotic teamwork – where multiple robot ‘minds’ work together to solve problems, pooling the input they get from their surroundings and searching through networked databases of possible solutions, based on elements from the fields of artificial intelligence, computer science, and social and cognitive psychology.
He also works with a company called CogniTeam, headed by Dr. Yehuda Elmaliach, a former student of his, which brings Kaminka’s theories and ideas to life. Among the company’s products is CogniTao, which allows robots to “think as one;” a plug-and-play system that out of the box lets groups of robots make decisions based on artificial intelligence programming. Other products and services enable robots to navigate and map routes for vehicles, intelligently patrol areas and alert humans in case of a problem, enable “robot vision” (where a robot pushes ahead into uncharted territory and reports back on conditions), and even building “artificial humans” – models that react like people do in specific situations (like car crashes), in order to develop better ways to deal with them.
Among CogniTeam’s clients are Israel Aeorspace Industries, The Ministry of Defense (for its MAFAT indoor robot swarms project), the IDF, and Bar Ilan itself, which has Israel’s largest robotics research group.
Many of the projects commercialized at CogniTeam have their roots in the research Kaminka and his students conducted at Bar Ilan. He is one of few researchers in the world who studies human-robot interface in groups, research which focuses on the operation of teams of robots by a single operator. For this project he and his students have been developing new algorithms which allow robots to operate autonomously, and share information selectively with the human operator, alongside innovative user interfaces for robotic systems.
Kaminka’s research has also had implications beyond robotics, contributing to the scientific study of intelligent agents and multi-agent systems, composed of intelligent software entities that coordinate and collaborate with each other in “swarms,” using artificial intelligence. In addition, he has conducted research into simulations of virtual entities, moving and behaving as complex individuals, collaborative teams, and crowds.
Beyond his academic and commercial activities, Kaminka, together with Dr. Eli Kolberg of Bar-Ilan’s Faculty of Engineering, of the first Israeli team participating in the RoboCup games, the premier league for soccer-playing robots, and has led Israel to the ranks of the world’s top 16 teams.
Summing up his work, Kaminka said that “even given the right mechanics and electronics, robots cannot interact effectively with others, unless their computational brains are made to think about others. My goal is to understand social intelligence; to understand the transition from a single mind to many; to build robots that are socially-intelligent; that are able to reason about, manipulate, collaborate with, and coordinate with other robots and humans; and to build computational models that explain social intelligence, that allow replication of it, that facilitate predictions of its occurrence, and that enable measurement and quantification.”