Autonomous agents that sense the world, make decisions, and actuate
Imagine multiple intelligent robots that have different sets of complementary capabilities and the ability to proactively invoke help from external resources, such as the web and humans thereby providing a seamless synergy of intelligence. Dr. Manuela Veloso, of Carnegie Mellon University, is working on just that as she researchers artificial intelligence and robotics. She hopes to one day produce the effective construction of autonomous agents where cognition, perception, and action are combined to address planning, execution, and learning tasks. With her students, Dr. Veloso investigates effective planning, execution, and learning algorithms for autonomous robots and agents within her CORAL laboratory. Such artificial intelligence and robotics will lead to the automation and assistance of many needs in society including working alongside patients in hospitals and nursing homes, assisting with secretarial work, and many other applications that will service humans.
As a full professor, with more than twenty years of research, Dr. Veloso’s expertise in artificial intelligence and robotics is likely to lead to innovative technologies. Her autonomous robots are perhaps most incredible because rather than being completely independent, they are capable of asking for help when they need it. With the understanding that robots will always have limitations compared to humans, Dr. Veloso’s approach to intelligence has been to create a team of robots that can complement one another to complete important tasks. In addition, her robots are capable of learning by optimizing how to perform actions after they have completed them. Dr. Veloso’s innovative approach to robotics has helped to craft some of the most successful indoor robots in existence. In fact, in November of 2014, her robots successfully reached the 620 mile marker with no incidents or accidents.
Current Research includes:
-
Perceiving the World: Robots are equipped with various types of sensors in order to perceive the world and understand the theme or situation going on around them. Dr. Veloso is working to make the robots more capable of perceiving the world accurately to assess specific situations.
-
Making Decisions: Dr. Veloso’s hopes to make robots able to plan and make decisions towards the accomplishments of such goals. Therefore, her robots are able to make plans that benefit humans.
-
Interacting with Humans: If robots are indoors, they need to interact with humans. To accomplish this, Dr. Veloso works on the interaction problem between humans and robots.
-
Learning from Experience: Dr. Veloso works to make robots better at whatever they are doing through experience. Therefore, the robots are able to improve upon action-based tasks as they continue to interact with the world.
Bio
Professor Veloso has always liked math in some sense. She completed a M.Sc. in electrical engineering in Portugal and her subsequent masters involved using computers to make automated company databases. When she finished her Ph.D. in 1992, she moved from theoretical to applied research. Since 1992, she has been working solely on robotics which has been an excellent platform for assisting the world.
Aside from research, as a native of Portugal, she loves the ocean and everything related to it. She enjoys traveling and has enjoyed spending time in many different countries. In addition, she speaks five languages fluently. As an active mother of two, she enjoys playing squash and reading.
Website: www.cs.cmu.edu/~mmv