The $500 hobby-shop helicopter whirling through a laboratory in the basement of Homewood's Barton Hall is not there to amuse the undergraduates. The copter is a key teaching tool in a demanding electrical engineering course aimed at preparing students for high-tech jobs in the aviation and automotive industries. Students in this class must program a computer to guide the aircraft safely through various maneuvers without help from a human "pilot." They cannot use the remote control box that normally steers such helicopters.
This aerial challenge has become part of a long-standing course called Control Systems Design, or Electrical and Computer Engineering 454. This spring, students for the first time will use model helicopters to see if the commands they've devised for a computer mockup will run a real-life flying machine. "The students will first do a simulation on the computer," explains Pablo Iglesias, an associate professor in the Whiting School's Department of Computer and Electrical Engineering. "Then, once they've convinced me that their computer program can fly the helicopter without wrecking it, I'll let them hook it up to the model. But in this class, it's important to remember that the emphasis is not on running the helicopters. The emphasis is on learning the basic principles of designing control systems."
Control systems are automatic devices that direct a machine to complete certain tasks. Examples include a plane's autopilot or a car's cruise control. In the Hopkins course, small teams of students must design a program that, once activated, will cause a model helicopter to rise a certain distance above the lab floor and then hover at that location for a specified period of time. Using sensors, the helicopter must be able to automatically stabilize itself after being "bumped" in a way that mimics the effect of wind gusts on a real aircraft. The students' performance on these projects accounts for 30 percent of their grade.
By using model copters, Iglesias is spicing up a course that in the past relied almost solely on computer simulations. When he took over the Control Systems class about five years ago, he found lab equipment that had little relation to real-world machines and a course outline that left students undermotivated. Since then, Iglesias has livened up the course by giving students a taste of challenges they someday may face in the working world. For example, Iglesias, who has collaborated on "intelligent highway" projects, has instructed his students to design a control system that could direct a car to automatically and smoothly speed up and pass another vehicle without intervention by the driver.
Until this year, students could only test such projects through computer simulations. But advances in technology now permit students to see how well their homemade control systems work on real-life machines, such as the model helicopters. "We now have software that finally allows us to put hardware in the loop in these projects," Iglesias says. "We can replace a part of the computer simulation with a second computer that actually controls the hardware and causes it to fly inside the lab. For these students, the computer system that controlled the simulated aircraft can now be used to control the hobby-shop helicopter."
For safety reasons, the model helicopters used in the lab remain attached to a swiveling metal stand. But even with this restraining device, the copters are capable of ascending, descending and flying from side to side like a real aircraft. For teaching purposes, the model helicopters are hardwired to the computer that sends commands to them. But the system also could be set up as a wireless operation, Iglesias says.
Iglesias' course is primarily for seniors majoring in electrical and computer engineering, biomedical engineering and mechanical engineering. Although graduate-level engineering programs on other campuses allow students to work with autonomous vehicles such as the model helicopters, undergraduates very rarely get the chance to do so, Iglesias says. "I wanted to show the students that computer simulations are not entirely realistic," he says. "It's only when you work with the hardware itself that you find out there are unforeseen limits to what you can do in the real world."