Featured

2019/03/21 Using AI to Build Better Human-Machine Teams

The inability of artificial intelligence (AI) to represent and model human partners is the single biggest challenge preventing effective human-machine teaming today. Current AI agents are able to respond to commands and follow through on instructions that are within their training, but are unable to understand intentions, expectations, emotions, and other aspects of social intelligence that are inherent to their human counterparts. This lack of understanding stymies efforts to create safe, efficient, and productive human-machine collaboration. “As humans, we are able to infer unobservable states, such as situational beliefs and goals, and use those to predict the subsequent actions, reactions, or needs of another individual,” said Dr. Joshua Elliott, a program manager in DARPA’s Informatio...

2019/03/21 Designing Chips for Real Time Machine Learning

The current generation of machine learning (ML) systems would not have been possible without significant computing advances made over the past few decades. The development of the graphics-processing unit (GPU) was critical to the advancement of ML as it provided new levels of compute power needed for ML systems to process and train on large data sets. As the field of artificial intelligence looks towards advancing beyond today’s ML capabilities, pushing into the realms of “learning” in real-time, new levels of computing are required. Highly specialized Application-Specific Integrated Circuits (ASICs) show promise in meeting the physical size, weight, and power (SWaP) requirements of advanced ML applications, such as autonomous systems and 5G. However, the high cost of design and implementa...

2019/03/19 DARPA Launches Social Media Platform to Accelerate R&D

DARPA aims to apply the power of social networks to research and development via a novel social media platform purpose-built to quicken the pace of U.S. technology development. Polyplexus, which was recently launched as a public platform, is designed to increase access to scientific evidence and accelerate hypothesis development, research proposal generation, and research sponsor engagement. The platform’s real-time interactions and functionality are designed to dramatically shorten the time between the emergence of concepts and the submission of worthy proposals. The platform facilitates connections among experts across academic disciplines so they can propel novel research opportunities together. Beta-launched for academics only in 2018, Polyplexus is now open to the broader researc...

2019/03/12 Progress on Lifelong Learning Machines Shows Potential for Bio-Inspired Algorithms

Today’s machine learning systems are restricted by their inability to continuously learn or adapt as they encounter new situations; their programs are fixed after training, leaving them unable to react to new, unforeseen circumstances once they are fielded. Adding new information to cover programming deficits overwrites the existing training set. With current technology, this requires taking the system offline and retraining it on a dataset that incorporates the new information. It is a long and arduous process that DARPA’s Lifelong Learning Machines (L2M) program is working to overcome. “The L2M program’s prime objective is to develop systems that can learn continuously during execution and become increasingly expert while performing tasks, are subject to safety limits, and capable of app...

2019/03/07 DARPA Seeks Tools to Capture Underground Worlds in 3D

DARPA is seeking information on state-of-the-art technologies and methodologies for advanced mapping and surveying in support of the agency’s Subterranean (SubT) Challenge. Georeferenced data – geographic coordinates tied to a map or image – could significantly improve the speed and accuracy of warfighters in time-sensitive active combat operations and disaster-related missions in the subterranean domain. Today, the majority of the underground environments are uncharted or inadequately mapped, including human-made tunnels, underground infrastructure, and natural cave networks. Through the Request for Information, DARPA is looking for innovative technologies to collect highly accurate and reproducible ground-truth data for subterranean environments, which would potentially disrupt and posit...

2019/03/01 DARPA Modernizes Small Business Innovation Research Program

Small Business Innovation Research and Small Business Technology Transfer (SBIR/STTR) are the principal set-aside programs for small business participation in federal research and development funding, yet the requirements for administering and managing these programs have not changed significantly in decades. To keep pace with discovery in science and technology worldwide, DARPA now intends to release SBIR/STTR opportunities on an out-of-cycle basis, separate from the three pre-determined announcements traditionally issued directly through the Department of Defense (DoD). The change is expected to reduce the overall time from opportunity announcement to contract award. Prior to the change, the timeline for SBIR/STTR funding opportunities was managed independently of DARPA’s primary technol...

2019/02/27 Taking the Next Step in Quantum Information Processing

Universal quantum computers with millions of quantum bits, or qubits – which can represent a one, a zero, or a coherent linear combination of one and zero – would revolutionize information processing for commercial and military applications. Realizing that vision, however, is still decades away. The problem is the performance and reliability of quantum devices depend on the length of time the underlying quantum states can remain coherent. If you wait long enough, interactions with the environment will make the state behave like a conventional classical system, removing any quantum advantage. Often, this coherence time is significantly short, which makes it difficult to perform any meaningful computations. To exploit quantum information processing before fully fault-tolerant quantum compute...

2019/02/19 A New Layer of Medical Preparedness to Combat Emerging Infectious Disease

DARPA has selected five teams of researchers to support PREventing EMerging Pathogenic Threats (PREEMPT), a 3.5-year program first announced in January 2018 to reinforce traditional medical preparedness by containing viral infectious diseases in animal reservoirs and insect vectors before they can threaten humans. Through studies in secure laboratories and simulated natural environments, the PREEMPT researchers will model how viruses might evolve within animal populations, and assess the safety and efficacy of potential interventions. Autonomous Therapeutics, Inc., Institut Pasteur, Montana State University, The Pirbright Institute, and the University of California, Davis, lead the PREEMPT teams. “DARPA challenges the PREEMPT research community to look far left on the emerging threat timel...

2019/02/15 Five Teams of Researchers Will Help DARPA Detect Undersea Activity by Analyzing Behaviors of Marine Organisms

Goliath grouper, black sea bass, and snapping shrimp, along with bioluminescent plankton and other microorganisms, are set to be the unlikely heroes of DARPA’s Persistent Aquatic Living Sensors (PALS) program. Five teams of researchers are developing new types of sensor systems that detect and record the behaviors of these marine organisms and interpret them to identify, characterize, and report on the presence of manned and unmanned underwater vehicles operating in strategic waters. This new, bio-centric PALS technology will augment the Department of Defense’s existing, hardware-based maritime monitoring systems and greatly extend the range, sensitivity, and lifetime of the military’s undersea surveillance capabilities. DARPA first announced the PALS program in February 2018 with the goal...

2019/02/14 Teaching AI Systems to Adapt to Dynamic Environments

Current AI systems excel at tasks defined by rigid rules – such as mastering the board games Go and chess with proficiency surpassing world-class human players. However, AI systems aren’t very good at adapting to constantly changing conditions commonly faced by troops in the real world – from reacting to an adversary’s surprise actions, to fluctuating weather, to operating in unfamiliar terrain. For AI systems to effectively partner with humans across a spectrum of military applications, intelligent machines need to graduate from closed-world problem solving within confined boundaries to open-world challenges characterized by fluid and novel situations. To attempt this leap, DARPA today announced the Science of Artificial Intelligence and Learning for Open-world Novelty (SAIL-ON) program. ...

2019/02/11 Architecting a New Breed of High Performance Computing for Virtual Training Environments

The testing, evaluation and training of future military systems will increasingly take place in virtual environments due to rising costs and system complexity as well as the limited availability of military ranges. Virtual simulators are already used to augment real-world training for modern fighter aircraft pilots, and they hold significant promise for addressing the rigorous demands of testing and training AI-enabled technologies. Current simulated environments, however, rely on conventional computing that is incapable of generating the computational throughput and speed to accurately replicate real-world interactions, model the scale of physical test ranges or meet the technical requirements of more complex systems. “Virtual environments could significantly aid the military by creating ...

2019/02/06 Intelligent Healing for Complex Wounds

Blast injuries, burns, and other wounds experienced by warfighters often catastrophically damage their bones, skin, and nerves, resulting in months to years of recovery for the most severe injuries and often returning imperfect results. This long and limited healing process means prolonged pain and hardship for the patient, and a drop in readiness for the military. However, DARPA believes that recent advances in biosensors, actuators, and artificial intelligence could be extended and integrated to dramatically improve tissue regeneration. To achieve this, the new Bioelectronics for Tissue Regeneration (BETR) program asks researchers to develop bioelectronics that closely track the progress of the wound and then stimulate healing processes in real time to optimize tissue repair and regenera...