Control Ambassadors

I t is not often that the future of control is discussed in the mainstream media, but it occurred in a column in the March 17, 2014 issue of Barron’s Technology Week [1]. The author, impressed with the advances in sensor technology displayed at the 2014 South by Southwest show in Austin, Texas, concludes the article with conjectures about future smart control devices that new sensors will enable. I was reminded of this article by conversations at the American Control Conference (ACC) in Portland, Oregon. Many types of sensors are becoming cheap and ubiquitous. An apt example is the typical smartphone, which contains an inertial measurement unit, a camera, a computer, satellite navigation, and communications devices sufficient to navigate and control autonomous vehicles. Sensor-stabilized quadrotors rapidly went from the research laboratory to commercially available toys. Vision-based intelligent cruise control and lane departure warning, enabled by cheap cameras, are now commercially available. The Barron’s column follows the ubiquitous sensor theme of [2] but neglects the huge advances that have occurred in computational hardware and algorithms; the speed, ubiquity, and reliability of communications; and advances in control and online data processing that have occurred in the last decades. Together, these advances enable a new era of control applications, including to large-scale, geographically distributed, dynamic systems. Discussions after the ACC special sessions identified some of the missing theoretical and practical tools needed to automatically organize and extract the essential information from the huge quantities of data and manipulate the many control variables effectively, reliably, and automatically. Cyberphysical systems research has actively considered related issues for such applications as the smart grid and smart cities, but many of the potential applications may be beyond our current imaginations. Sensing and communication ubiquity are affecting observability and controllability in fundamental ways for applications where performance is currently limited either by lack of distributed sensing or actuation. How will cell phones, as ubiquitous mobile traffic sensors, alter the field of transportation flow control? Control Ambassadors