Dr. Daniel Fischer
Senior Engineer
Kinectrics
Toronto, Ontario, Canada

Thursday, September 11, 2008, 6:00 p.m. – 8:00 p.m.
*** Refreshments will be served ***

Many of us have heard of Fuzzy Logic and Neural Networks, two major Artificial Intelligence domains. Perhaps, when looking at Neural Networks, we were exposed to different network structures (e.g. feed forward with backpropagation learning, radial basis function, etc), or different learning approaches (supervised vs. unsupervised). When considering Fuzzy Logic, we have learned about fuzzy sets, fuzzy membership functions, fuzzification, inference and aggregation, defuzzification. All this information is useful in order to understand the details of what goes on inside a Neural Network or a Fuzzy System when they perform their computations. However, the same details may not be particularly helpful in showing when a certain technology is likely to be successful when applied to an application. In this tutorial, as an application example, we will show how Artificial Intelligence can be used to achieve a task we, humans, are pretty good at: stopping a car before it crashes into a wall. We will use a Neural Network to implement our acquired driving experience: the ability to estimate the stopping distance, given the vehicle’s speed and applied pressure on the break pedal. We will use a Fuzzy Logic system to capture the driver’s style: aggressive, delaying pressing on the break pedal until the last moment, or relaxed, starting the stopping process early. We will show how we can implement a smooth stop and will implement all this in a simulation running under Matlab Simulink. It is unlikely that this tutorial will improve our driving abilities, however it is hoped that our understanding of applied Artificial Intelligence would increase.

Dr. Daniel Fischer is currently a Senior Engineer in the Transmission and Distribution Technologies business at Kinectrics, Ontario. Over the last two decades, he contributed to the design and implementation of advanced signal processing techniques that improved the monitoring and performance of the transmission and distribution networks. Dr. Fisher successfully applied modern Artificial Intelligence methodologies to a wide array of failure detection systems responsible for detecting underground power cable oil leaks, generator overheating, and hydro dam structure degradation. Dr. Fisher has extensive experience in both hardware and software design of systems having strong real time requirements. Dr. Fisher has a B.A.Sc. in Engineering Science and a M.A.Sc. in Electrical Engineering both from the University of Toronto, and a Ph.D. in Electrical and Computer Engineering from McMaster University. He is a senior member of the IEEE and a registered Professional Engineer in the Province of Ontario.