Our Robots, Ourselves: Robotics and the Myths of Autonomy [Bookshelf]

This book focuses on the autonomy of robots and our technological inclination toward constructing machines capable of navigating among us better than—or at least as well as—ourselves. The author reveals his skepticism with this view. He has considerable experience in the foundational stages of autonomous engineering systems and, with his historian’s eye for evidence, is able to develop unifying themes from his observation of people interacting with these systems. Invariably, the replacement of humans in the loop winds up enhancing the performance of the human operators, since the decision support needed for autonomy improves the information available to the operator, without the attendant need for blind trust of the machine. The author explores examples from submersible systems, lunar landers, commercial aircraft, and military drones. Throughout the book, the author concentrates on the interplay between humans and machines and, from this viewpoint, develops his theses. The book is a very good read. The arguments are evidence based and cogent. The book provides a thought-provoking examination of autonomy in robotics that cracks opens a few central technological and human questions. As control engineers, it is very accessible and familiar in scope. More importantly, it forces us to examine the headlong rush to autonomy in systems and to ask whether this is a fundamentally flawed endeavor. Control will play a central role in the realization of many aspects of these systems, so it is good that we are goaded into a critical examination of their purpose.