Back to news

February 4, 2019 | International, C4ISR

DARPA: Building Trusted Human-Machine Partnerships

A key ingredient in effective teams – whether athletic, business, or military – is trust, which is based in part on mutual understanding of team members' competence to fulfill assigned roles. When it comes to forming effective teams of humans and autonomous systems, humans need timely and accurate insights about their machine partners' skills, experience, and reliability to trust them in dynamic environments. At present, autonomous systems cannot provide real-time feedback when changing conditions such as weather or lighting cause their competency to fluctuate. The machines' lack of awareness of their own competence and their inability to communicate it to their human partners reduces trust and undermines team effectiveness.

To help transform machines from simple tools to trusted partners, DARPA today announced the Competency-Aware Machine Learning (CAML) program. CAML aims to develop machine learning systems that continuously assess their own performance in time-critical, dynamic situations and communicate that information to human team-members in an easily understood format.

“If the machine can say, ‘I do well in these conditions, but I don't have a lot of experience in those conditions,' that will allow a better human-machine teaming,” said Jiangying Zhou, a program manager in DARPA's Defense Sciences Office. “The partner then can make a more informed choice.”

That dynamic would support a force-multiplying effect, since the human would know the capabilities of his or her machine partners at all times and could employ them efficiently and effectively.

In contrast, Zhou noted the challenge with state-of-the-art autonomous systems, which cannot assess or communicate their competence in rapidly changing situations.

“Under what conditions do you let the machine do its job? Under what conditions should you put supervision on it? Which assets, or combination of assets, are best for your task? These are the kinds of questions CAML systems would be able to answer,” she said.

Using a simplified example involving autonomous car technology, Zhou described how valuable CAML technology could be to a rider trying to decide which of two self-driving vehicles would be better suited for driving at night in the rain. The first vehicle might communicate that at night in the rain it knows if it is seeing a person or an inanimate object with 90 percent accuracy, and that it has completed the task more than 1,000 times. The second vehicle might communicate that it can distinguish between a person and an inanimate object at night in the rain with 99 percent accuracy, but has performed the task less than 100 times. Equipped with this information, the rider could make an informed decision about which vehicle to use.

DARPA has scheduled a pre-recorded webcast CAML Proposers Day for potential proposers on February 20, 2019. Details are available at: https://go.usa.gov/xE9aQ.

The CAML program seeks expertise in machine learning, artificial intelligence, pattern recognition, knowledge representation and reasoning, autonomous system modeling, human-machine interface, and cognitive computing. To maximize the pool of innovative proposal concepts, DARPA strongly encourages participation by non-traditional proposers, including small businesses, academic and research institutions, and first-time Government contractors.

DARPA anticipates posting a CAML Broad Agency Announcement solicitation to the Federal Business Opportunities website in mid-February 2019.

https://www.darpa.mil/news-events/2019-01-31

On the same subject

  • The Army's M1 Abrams Tank Is About To Get Even Deadlier

    January 8, 2019 | International, Land, C4ISR

    The Army's M1 Abrams Tank Is About To Get Even Deadlier

    by Kris Osborn The Army is engineering new AI-enabled Hostile Fire Detection sensors for its fleet of armored combat vehicles to identify, track and target incoming enemy small arms fire. This system, integrated onto Apache Attack helicopters, uses infrared sensors to ID a “muzzle flash” or heat signature from an enemy weapon. The location of enemy fire could then be determined by a gateway processor on board the helicopter able to quickly geolocate the attack. The Army is engineering new AI-enabled Hostile Fire Detection sensors for its fleet of armored combat vehicles to identify, track and target incoming enemy small arms fire. Even if the enemy rounds being fired are from small arms fire and not necessarily an urgent or immediate threat to heavily armored combat vehicles such as an Abrams, Stryker or Bradley, there is naturally great value in quickly finding the location of incoming enemy small arms attacks, Army weapons developers explain. There are a range of sensors now being explored by Army developers; infrared sensors, for example, are designed to identify the “heat” signature emerging from enemy fire and, over the years, the Army has also used focal plane array detection technology as well as acoustic sensors. “We are collecting threat signature data and assessing sensors and algorithm performance,” Gene Klager, Deputy Director, Ground Combat Systems Division, Night Vision and Electronic Sensors Directorate, told Warrior Maven in an interview last year. Klager's unit, which works closely with Army acquisition to identify and at times fast-track technology to war, is part of the Army's Communications, Electronics, Research, Development and Engineering Center (CERDEC). Army senior leaders also told Warrior Maven the service will be further integrating HFD sensors this year, in preparation for more formals testing to follow in 2019. Enabling counterattack is a fundamental element of this, because being able to ID enemy fire would enable vehicle crews to attack targets from beneath the protection of an armored hatch. The Army currently deploys a targeting and attack system called Common Remotely Operated Weapons System, or CROWS; using a display screen, targeting sensors and controls operating externally mounted weapons, CROWS enables soldiers to attack from beneath the protection of armor. “If we get a hostile fire detection, the CROWS could be slued to that location to engage what we call slue to cue,” Klager said. Much of the emerging technology tied to these sensors can be understood in the context of artificial intelligence, or AI. Computer automation, using advanced algorithms and various forms of analytics, can quickly process incoming sensor data to ID a hostile fire signature. “AI also takes other information into account and helps reduce false alarms,” Klager explained. AI developers often explain that computer are able to much more efficiently organize information and perform key procedural functions such as performing checklists or identifying points of relevance; however, many of those same experts also add that human cognition, as something uniquely suited to solving dynamic problems and weighing multiple variables in real time, is nonetheless something still indispensable to most combat operations. Over the years, there have been a handful of small arms detection technologies tested and incorporated into helicopters; one of them, which first emerged as something the Army was evaluating in 2010 is called Ground Fire Acquisition System, or GFAS. This system, integrated onto Apache Attack helicopters, uses infrared sensors to ID a “muzzle flash” or heat signature from an enemy weapon. The location of enemy fire could then be determined by a gateway processor on board the helicopter able to quickly geolocate the attack. While Klager said there are, without question, similarities between air-combat HFD technologies and those emerging for ground combat vehicles, he did point to some distinct differences. “From ground to ground, you have a lot more moving objects,” he said. Potential integration between HFD and Active Protection Systems is also part of the calculus, Klager explained. APS technology, now being assessed on Army Abrams tanks, Bradleys and Strykers, uses sensors, fire control technology and interceptors to ID and knock out incoming RPGs and ATGMs, among other things. While APS, in concept and application, involves threats larger or more substantial than things like small arms fire, there is great combat utility in synching APS to HFD. Full article: https://nationalinterest.org/blog/buzz/armys-m1-abrams-tank-about-get-even-deadlier-40847

  • PAE awarded ten-year contract to provide aircraft maintenance to the U.S Customs and Border Protection Agency

    May 20, 2020 | International, Aerospace, Security

    PAE awarded ten-year contract to provide aircraft maintenance to the U.S Customs and Border Protection Agency

    May 19, 2020 - Falls Church, Va. – PAE (NASDAQ: PAE, PAEWW), a global leader in delivering smart solutions to the U.S. government and its allies, was awarded the National Aviation Maintenance and Logistics Services contract by U.S. Customs and Border Protection, an agency of the U.S. Department of Homeland Security. The contract has a one-year base period of performance with nine one-year options, a potential three-month extension and an estimated total value of more than $1.26 billion. PAE President and CEO John Heller said the award recognizes PAE's dependability and specialized expertise for aircraft maintenance. “With our long track record serving CBP's aircraft maintenance needs, PAE has proven to be a trusted partner in supporting U.S. national security,” Heller said. “We look forward to continuing our support to this critical national security customer as we apply our innovative solutions supporting the CBP mission and fleet of over 200 aircraft.” “This is truly a team win and a tribute to the more than 650 dedicated men and women on this program,” said PAE Vice President of CBP Program Management Rob Ulses. “These hard-working individuals have established a real partnership with CBP that allows us not only to support day-to-day operations, but to adapt and react to special missions.” PAE will continue to provide safe and ready aircraft to ensure the U.S. government meets operational commitments to safeguard America's borders. PAE will provide this essential support from aviation operational sites at military bases, civilian airfields and alternate locations across the United States. The broad scope of work extends from scheduled and unscheduled aircraft maintenance and repair to managing fueling, logistics and supplies. About PAE For 65 years, PAE has tackled the world's toughest challenges to deliver agile and steadfast solutions to the U.S. government and its allies. With a global workforce of about 20,000 on all seven continents and in approximately 60 countries, PAE delivers a broad range of operational support services to meet the critical needs of our clients. Our headquarters is in Falls Church, Virginia. Find us online at pae.com, on Facebook, Twitter and LinkedIn. Forward-Looking Statements This press release may contain a number of “forward-looking statements” as defined in the Private Securities Litigation Reform Act of 1995. These forward-looking statements include, but are not limited to, statements about PAE's possible or assumed future results of operations, financial results, backlog, estimation of resources for contracts, strategy for and management of growth, needs for additional capital, risks related to U.S. government contracting generally, including congressional approval of appropriations, and contract delays or cancellations caused by competitors' bid protests of contract awards received by us. These forward-looking statements are based on PAE's management's current expectations, estimates, projections and beliefs, as well as a number of assumptions concerning future events. These forward-looking statements are not guarantees of future performance, conditions or results, and involve a number of known and unknown risks, uncertainties, assumptions and other important factors, many of which are outside PAE's management's control, that could cause actual results to differ materially from the results discussed in the forward-looking statements. Forward-looking statements included in this release speak only as of the date of this release. PAE does not undertake any obligation to update its forward-looking statements to reflect events or circumstances after the date of this release except as may be required by the federal securities laws. For media inquiries regarding PAE, contact: Terrence Nowlin Senior Communications Manager PAE 703-656-7423 terrence.nowlin@pae.com For investor inquiries regarding PAE, contact: Mark Zindler Vice President, Investor Relations PAE 703-717-6017 mark.zindler@pae.com View source version on PAE: https://www.pae.com/news/pae-awarded-ten-year-contract-provide-aircraft-maintenance-us-customs-and-border-protection

  • Understanding Warfighter Performance from the Inside Out

    January 23, 2019 | International, Land

    Understanding Warfighter Performance from the Inside Out

    Measuring Biological Aptitude (MBA) aims to identify, understand, and monitor in real time the biology that underlies success in specialized roles A new program out of DARPA's Biological Technologies Office could help the Department of Defense enhance and sustain military readiness both by revolutionizing how troops train, perform, and recover, and by mitigating shortages of highly qualified candidates for extremely specialized roles. The anticipated outputs of the Measuring Biological Aptitude (MBA) program are a set of biomarkers — measurable indicators of biological processes — that correspond to traits of highly effective performance in a given role, along with new tools to measure and report on those biomarkers in real time. This information will enable individual warfighters to understand and affect the underlying biological processes that govern their success. MBA technologies could improve training, team formation, mission performance, and post-mission recovery, yielding a better prepared, more effective, more resilient force. At its core, MBA seeks to shed light on the biological factors and processes that support peak performance in each of a set of military specializations. The research will work backwards from phenotypes — that is, how an individual's fixed genetic code expresses as externally observable cognitive, behavioral, or physical traits — and attempt to establish the biological mechanisms that translate underlying genetic makeup into phenotypic traits. At present, those mechanisms of translation — also known as expression circuits — are largely a mystery. MBA researchers will develop new assays and technologies to monitor and report on the biomarkers that reveal the activity of key expression circuits. “With existing technology scientists are able to read out genotype and measure and observe certain aspects of phenotype. Most of what happens in between is a black box,” said Eric Van Gieson, the MBA program manager. “DARPA believes that the information inside the box — these expression circuits — can be predictive of how an individual will respond to a given stimulus or scenario, and more importantly, we believe it will help inform the individual on how to improve their performance throughout their career.” Researchers supporting MBA will initially analyze samples and other data collected from high-performing troops across select military specializations to identify biological signatures of successful performance in each of those roles and determine how they can be measured. For instance, maintaining a lowered heart rate during combat is a valuable trait and easily measured with existing wearable technology. Adaptable problem solving, resilience, and cognitive flexibility are extremely valuable, but less easily measured. MBA analyses should reveal an array of such traits and the expression circuits responsible for them. If DARPA succeeds, the resulting MBA system could support military readiness in various ways. The first improvement relates to how the military initially evaluates recruits and subsequently develops candidates for specialized roles. Many of these roles currently suffer from shortages of qualified candidates, even as more pervasive use of complex technologies and an expanding set of mission profiles are increasing demand for uniquely skilled personnel. For at least the past 50 years, initial assessment of military service members has remained essentially unchanged, comprising a basic medical screening, a standardized physical readiness test, and a written test known as the Armed Services Vocational Aptitude Battery (ASVAB) for enlisted personnel. Scores on the ASVAB feed into the preliminary determination of an individual's qualification for certain military occupational specialties. As a service member's career advances, future placement into other roles does not follow a prescribed protocol and can be based in large part on subjective measures. Against this backdrop, MBA technology could increase the objectivity of the criteria used by military selection committees, remove biases, and raise the baseline of performance for incoming recruits. Additionally, by taking biology into account, the results from MBA measurements could reveal to individuals career options that might not be apparent based on commonly accepted, externally observable traits alone. The second improvement ties to training, both before and after an individual pursues a military career. MBA technology could allow a user to assess his or her personal potential for specialized roles and proactively nurture the traits that are characteristic of successful performers. “Genotypes are fixed, but phenotypes are not. Biology is fundamentally adaptable, and that is the key to enabling performance improvements,” Van Gieson said. “What we're planning to deliver with MBA is a set of continuously updated information that empowers individuals to track their progress throughout their careers and quickly identify what aspects of training and preparation are the most productive.” Third, during missions commanders could employ real-time reporting of changes in service members' biomarkers to inform how a military operation unfolds, adding a layer of biological awareness to provide a more complete assessment of the mission space. Commanders could shift resources or adjust strategies and tactics based on how squad members are performing. Following a mission, biomarker reporting could likewise guide recovery practices and indicate potential health issues. The overall MBA program will be informed by consultations with independent expert advisors in the ethical, legal, social, and regulatory aspects of the work, with particular emphases on privacy, data protection, and responsible utilization of data by individuals. MBA performer teams will be required to provide medical guidance as part of any human study through an embedded genetic counselor, sports therapist, or similar specialist. “Human beings are extremely complex, and although we expect to gain valuable new insights by measuring biology, we also understand that people are not locked into predetermined fates,” Van Gieson explained. “Any breakthroughs we achieve in the MBA program will necessarily be used to address shortages in critical roles by expanding opportunities, not limiting them. If we can provide people with information on their unique biology, and empower them to affect and measure gains in key traits, we'll have opened career pathways that they may not have previously considered.” DARPA will hold a Proposers Day on February 12, 2019, in Arlington, Virginia, to provide more information about MBA and answer questions from potential proposers. For details of the event, visit https://go.usa.gov/xEZeT. A forthcoming Broad Agency Announcement will include complete program objectives, schedules, and metrics. Team should have experience in human performance, phenotyping, multi-scale biology, physiology, biomarker detection and tracking, device development, and various other aspects that will be specified in the announcement. https://www.darpa.mil/news-events/2019-01-22a

All news