The U.S. Army has made it clear that it wants smart robots capable of both working closely alongside soldiers and working autonomously. These unmanned systems must also adapt in real time to changing situations, conditions, and environments. Further, they should learn from experience.
A pipe dream? Not necessarily. Such robots do not exist today except in movies, but that is not discouraging the Army Research Laboratory (ARL) or Jon Bornstein, the Army’s collaborative alliance manager. Bornstein has been charged with developing the basic technology to make this robotic vision a reality. To that end, he launched the current Robotics Collaborative Technology Alliance (RCTA), a $60-plus million, five-year effort, which kicked off in April 2010. It is part of the ARL’s Autonomous Systems Enterprise that combines ARL’s internal research efforts with external research. Already the plan calls for a follow-on RCTA effort for the same or more money in five years.
The bottom line: There is a potential $120-plus million in funding for robotics companies and research institutions that are ready and able to engage in a collaborative effort on autonomous systems as a member of the RCTA consortium assembled by Bornstein.
Latest RCTA Consortium
In April, ARL directed the latest RCTA consortium, led by General Dynamics Robotic Systems (GDRS), to take on its current robotics challenge. The other partners in the RCTA include Boston Dynamics, Carnegie Mellon University, California Institute of Technology Jet Propulsion Laboratory, Florida A&M University, QinetiQ North America, University of Central Florida, and University of Pennsylvania. The participants are not expected to provide robotic products or robotic platforms. Rather, Bornstein seeks advanced technology that will enable future systems.
ARL uses cooperative agreements with partners to bring together diverse consortiums to develop and execute research plans that share financial, intellectual, personnel, and infrastructure resources from both the government and the private sector. The latest agreement is the third robotics-centered Collaborative Technology Alliance (CTA) initiated by ARL.
Central to the award is the idea of the cooperative agreement (not to be confused with a contract). A cooperative agreement enables the Army to enter into an arrangement for joint research and education. The Army intends to work collaboratively with the consortium members and expects a free, multidirectional flow of information and ideas among participants. A contract does not allow for such collaborative effort. However, should technology emerge from this effort that addresses a specific Army need, the work could transition into a procurement contract.
Although the current consortium is already formed and has been engaged since April, there remains an opportunity for latecomers to get involved. The Army is permitted to withhold up to 10 percent of the funding for new ideas that may crop up during the five-year period of the RCTA. In that case, the company takes its idea to the consortium and works as a subcontractor.
The previous RCTA focused on Laser Detection and Ranging (LADAR). At that time, ARL considered autonomous navigation a considerable challenge, primarily because of the difficulty in describing the environment in a way that captures the variability of natural environments. That RCTA focused on the problem of extracting the ground terrain surface from sparse 3D data using LADAR mobility sensors.
Bornstein has charged the latest RCTA with a different mission. As noted in Army documents: This alliance will bring together government, industrial, and academic institutions to address some of the fundamental scientific and technological underpinnings to enable the future deployment of highly autonomous unmanned systems, including ground, air, and surface vehicles, on the battlefield. The technical areas of interest include: (1) perception, (2) intelligence, (3) human-robot interaction, (4) dexterous manipulation and unique mobility, and (5) integrated performance.
As Bornstein sees it, these are areas now crucial to the advancement of future unmanned systems possessing a significant level of autonomy. The Army wants highly intelligent systems that work closely with soldiers and also work on their own.
The RCTA will focus on cutting-edge research in perception, intelligence, human-robot interaction, manipulation, and mobility. Since no single organization is likely to be capable of addressing all these areas, the Army seeks out a consortium. To be considered for the next RCTA, an organization must submit a proposal to the ARL. To get into the current consortium, the organization must submit a proposal to the current consortium.
In other documents, Bornstein elaborates on how the RCTA’s focus areas translate in practice. Specifically, the robot must act like another soldier in the squad. That means understanding the commander’s intent. It also must be highly capable while minimizing communication and reducing the need for soldier interaction. Of course, like a soldier, it must be flexible, robust, and reliable; able to adapt fully to new and different tactical and environmental conditions; and effectively operate in mixed environments. Also like a soldier, it should be able to learn from experience, maneuver unfettered in complex terrain, and be able to function in a world designed for humans. So it must be able to grasp small objects, open doors, or carry the wounded. What it comes down to is a robot with human-like characteristics.
Unique mobility, for example, presents a case in point. Today robots gain mobility through wheels or tracks. But what if the robot could crawl or slither like a snake? Such a robot might be ideal for penetrating dangerous tunnels, a common situation encountered in modern warfare. The Army would greatly prefer to send a slithering robot into such a situation than a live soldier.
The current RCTA award runs five years. By the fourth year, the ARL reviews the accomplishments with an eye toward its impact on robotics science and on the soldier. If the accomplishments warrant it, another five-year award can be initiated. With the latest RCTA awarded in 2010, expect the next one to begin ramping up in 2014. Before then, look for possible consortium subcontract opportunities.
|The Bottom Line