An AI expert goes to Arlington

Hector Muñoz-Avila, associate professor of computer science and engineering, has applied his expertise in artificial intelligence (AI) to a broad variety of endeavors in the last few years.

Using AI and sensors, Muñoz-Avila is helping structural engineers predict when bridges will fail. Using the principles of goal-driven autonomy, a field in which he is a pioneer, he is helping electrical engineers orchestrate the components needed to generate energy from ocean waves.

In a collaboration with psychologists, Muñoz-Avila is developing computational models based on neural networks to represent semantic interference and other phenomena that occur inside the brain.
 
Recently, Muñoz-Avila was asked by NSF to serve as a program director in its Directorate for Computer and Information Science and Engineering in Arlington, Va. Last fall, he began a two-year rotation in the Robust Intelligence Cluster of the directorate’s Division of Information and Intelligent Systems (IIS).

NSF program directors make recommendations about proposal funding. They help shape the nation’s direction in science, engineering and education and in the defining of new funding opportunities.

The NSF appointment comes on the heels of a sabbatical Muñoz-Avila spent at the Norwegian University of Science and Technology (NTNU) in Trondheim, where he studied the application of AI to deep-sea oil-drilling rigs. The goal of the project is to avoid expensive shutdowns by predicting when physical systems might fail.

Muñoz-Avila has chaired several international scientific meetings, including the Sixth International Conference on Case-Based Reasoning (ICCBR-05) and the twenty-fifth Innovative Applications of AI Conference (IAAI-13).

One common theme to his varied research activities is the application of case-based reasoning and other AI techniques to intelligent systems, or agents, that have the ability to sift through thousands of stimuli and data points, and to pinpoint and correct unusual patterns or anomalies. These agents can be robots, automated computer game players or systems that monitor an electrical grid. Their ability to learn from their experiences and mistakes and to take corrective action without human intervention is part of the new field of goal-driven autonomy.

In 2012, Muñoz-Avila received a three-year research grant from NSF to develop autonomous agents that dynamically identify and self-select their goals, and to test these agents in computer games.