Virginia Tech

Virginia Tech is a public university located in Blacksburg, Virginia. Virginia Tech maintains the Virginia Center for Autonomous Systems (VaCAS), a branch of College of Engineering focused on the research and development of autonomous systems technology. VaCAS prides itself on being diverse in its studies of autonomous systems, researching the use of autonomous systems on land, in the air, in the sea, and in space, as well as everything between control theory and practical applications in daily life.

Air
VaCAS's premier aerial autonomous system technology is the small platform for autonomous aerial research operations (SPAARO). Designed by master's student Justin Murtha, the SPAARO was designed with reliability in mind. According to a 2005 study by the Department of Defense, UAVs are ten times more likely to experience a technical failure than a manned aircraft. Craig Woolsey of the Nonlinear Systems Lab noted how, when it comes to small unmanned aircraft, companies are much less likely to spend time developing reliability for their systems, because the loss encountered by a small unreliable unmanned system is nothing compared to the loss encountered by an unreliable commerical air vehicle, which has the potential to cost not only an excessive sum of money, but many human lives.

Under contract from the United States Naval Air Systems Command (NAVAIR), Murtha set out to determine a cost effective method to make small UAVs reliable. The end result was a method of interpreting failure data to determine whether it would be more cost-effective in relation to reliability to purchase cheap replacement parts that wouldn't last as long or to purchase more expensive replacement parts that would last longer. Prominently, Murtha found a "knee" in the cost curve, or the point in which there is a diminishing return for the investment put into it. This research has resulted in the creation of the SPAARO, first built as a side project by Murtha. Before the SPAARO, Virginia Tech had been using Sig® Rascal airframes, modified for various research objectives, including carrying payloads. The issue that Murtha noted with these, however, was that they were being flown at nearly twice the weight they had been made to support, due to the structure of the airframes not being designed to support more than was already on it. The Sig® Rascal airframes were proving themselves to be obsolete, hence Murtha wanted his SPAARO to be able to withstand prolonged use with swapping out payloads and other modifications, and, crucially, to survive being flown by those without much experience in flying small UAVs. The SPAARO was supposed to become the basis for years of experiments to come, to act as a workhorse.

Land
In collaboration with TORC, a company with roots in Virginia Tech, students with VaCAS work on several land vehicles, with the aim of creating autonomous land vehicles. One such project created a Ford hybrid Escape that was able to navigate sixty miles in the city autonomously. This vehicle won third place in the 2007 DARPA Urban Challenge. Another similar project had researchers developing a two-vehicle system for the military, designed to scan for improvised explosive devices (IED) on planned routes. This system was called the autonomous remote-controlled HMMWVs (ARCH). It consists of one autonomous vehicle, also able to be controlled by remote control, followed by a manned chase vehicle. A similar system is being devised with Polaris ATVs to allow for greater application of the technology in more varied environments.

Water
Communication underwater is an issue. The usual channels used on land or in the air don't work underwater, making it difficult to communicate beneath the surface. A team of VaCAS researchers are working to create a viable communication method for autonomous underwater vehicles (AUV) to enable them to complete tasks autonomously, but while also communicating as little as necessary. To conduct this research, the team is using two of Virginia Tech's 475 AUVs. One test conducted was to have the two AUVs work together to locate the source of a sound underwater, a goal accomplished by the transmision of only seven one-way data packets.

Many people are involved in VaCAS's autonomous systems technologies relating to water. Dan Stilwell, of electrical and computer engineering, is in charge of algorithm development and field trials. Aerospace and ocean engineering is led by Wayne Neu. Mike Roan provides advice on sensor specifications. The entire project is funded by the Office of Naval research. While the 475 AUV is the workhorse for VaCAS's underwater testing, the team is working on developing a new AUV platform. This one under development is supposed to be capable of deploying an anchor and mooring itself on the ocean floor, to remain there for an extended period of time. It is supposed to travel all the way to the desired location, moor itself, and return entirely autonomously. It has currently been successfully demonstrated as a prototype, and a larger version of it is being developed for deployment by the Navy in coming years.

Above the surface of the water, VaCAS has another project of unmanned surface vehicles (USV). This fleet of inflatable boats with onboard computer systems is being developed for military purposes. Currently, military forces that are navigating a river or other body of water controlled by the opposition are at a tactical disadvantage. There is no telling what is further down the river or around the bend, or even what is below the surface. This is where VaCAS's USVs are coming into play; they are to be the scouts that determine the safety of a river.

The USVs must be able to navigate autonomously to be of any use, not to mention being able to build maps from high resolution imagery and provide information on any possible obstacles encountered. This is a task easier said than done, considering the wealth of hazards present in a river, ranging from fallen trees, jutting rocks, vehicle wreckage, and water depths. Speed is also a must; if a vehicle is in unfriendly territory, speed is its greatest defense, and as such, must be able to detect threats not only below the surface, but above the surface, and be able to act upon those threats within seconds. To reduce the load on the USV, it has been considered to also deploy various other autonomous vehicles under the water, on land, and in the air, yet, by doing so, it increases the computational load on the USV. The USV project is a joint project between Virginia Tech and the US Navy. Virginia Tech has provided an inflatable boat for increased portability and a system that gives the ability of surface scanning, while the Naval Postgraduate School has contributed a sonar-equipped boat to detect underwater obstacles. Virginia Tech's surface scanning technology relies on line-of-sight laser technology, and then combines gathered data about the river surface with incomplete map data and aerial photographs to provide an accurate, up to date map. This project was tested a success when a prototype was able to autonomously navigate four kilometers of river in twenty-five minutes, marking hazards that hadn't been noticed in a preliminary examination of the river.

Core Labs

 * The Autonomous Systems and Control Laboratory
 * The Center for Dynamic Systems Modeling and Control
 * The Mechatronics Laboratory
 * The Nonlinear Systems Laboratory
 * The Unmanned Systems Laboratory

Affiliate Labs

 * Center for Intelligent Materials, Systems, and Structures
 * Center for Human Computer Interaction
 * Collaborative Center for Multidisciplinary Sciences