MAE 5810 - Robot Perception

Fall. 3 credits. Letter grades only.

Prerequisite: ENGRD 2112, MATH 2940, and MATH 4710 (or ENGRD 2700 or ENGRD 2720); or graduate standing in a technical field. Co-meets with MAE 4810.

S. Ferrari.

An introductory course to robot perception techniques for modeling and planning heterogeneous and dynamic sensor measurements, and for processing the sensor feedback in the context of robot motions and environments. Methods for intelligent sensor fusion and robot perception in motion will be covered in detail in this course. Topics in artificial vision, acoustic propagation, and filtering will be discussed along with related algorithms inspired by neural networks, Bayesian networks, and information theory. Sensing problems and performance will be investigated in regard to benchmark problems, such as coverage, target search, target tracking, and treasure hunting, will be covered in-depth and demonstrated through applications drawn from environmental monitoring, sensing-and-pursuit games, surveillance, and human-robot interactions.

Outcome 1: Students will use methods for sensor fusion to inform robot perception.

Outcome 2: Students will implement artificial vision, acoustic propagation, and filtering along with neural networks, Bayesian algorithms, and information theory to identify robot states.

Outcome 3: Students will analyze benchmark problems such as coverage, target search, target tracking, and treasure hunting.