SBIR-STTR Award

Training is costly and time-intensive, yet absolutely essential in nearly every endeavor where human skills drive the enterprise. For example, training a fighter pilot to proficiency costs the USAF nearly $11 million. Ineffective training can result in loss of life, destruction of valuable equipment, and squandered person-hours. Training success can be quantified from the perspective of economics (minimizing cost-per-trainee), time (minimizing time-to-proficiency), person-power (maximizing throughput into the workforce), or performance metrics that vary by domain. Regardless of one’s perspective, the challenge is the same: how can training be optimized to accelerate learning and enhance retention? When instructors must use the same material for all learners, the risk is that some students will not quite grasp the material and are left without the necessary remediation as the content continues to progress on a set timeline. Conversely, more knowledgeable students may be bored and unable to achieve their full potential. Adaptive learning approaches were introduced to counter this “one-size-fits-all” approach and ensure that learners receive the content that is most appropriate for their current knowledge and skill level. However, recent advances in neurophysiological sensing technology have introduced new opportunities for even more rapid discoveries. “Neuroadaptive learning” employs a closed-loop brain-computer interface (BCI) to optimize training by adjusting itself in real time to each individual learner’s cognitive and affective state. Aptima, with our partners at Cognionics, Reflexion, and the University of Chicago, will incorporate cutting-edge cognitive and affective neurophysiological sensing technology, artificial intelligence and machine learning (AI/ML), and the Reflexion interactive learning environment, to develop this first-of-its kind solution. The NeuroAdapt system will integrate real-time training data from Reflexion with neurophysiological measures such as electroencephalography (EEG), heart-rate variability (HRV), galvanic skin response (GSR), and likely others such as functional near-infrared spectroscopy (fNIRS) and eye tracking. A powerful AI/ML engine based on Aptima’s existing platforms will adapt to each individual's performance and neurocognitive/affective state, learning the patterns and pathways that generate optimal learning outcomes, and adjusting the training paradigm to accelerate learning and enhan

Air Traffic Control (ATC) is a vital function for the US Air Force, ensuring the safety and efficiency of aircrew and ground personnel. ATC is cognitively demanding, sometimes likened to solving a puzzle: operators must actively track the three-dimensional position, speed, and attitude of multiple aircraft simultaneously, all while maintaining awareness of their destinations and objectives. This is a logistical problem requiring sharply honed cognitive skills in the domains of working memory, executive function, attention, spatial cognition, and stress management, as well as a social problem in which ATC operators must communicate effectively with pilots in their airspace to maintain collective situational awareness and ensure safe and efficient flow of traffic. Air Force ATC operators receive only 72 days of technical training. Therefore, even incremental improvements to effectiveness and efficiency of ATC training programs can have a significant impact; advanced solutions are needed to enhance and accelerate ATC training to maximize mission readiness and to reduce personnel and material costs. Adaptive learning methods are evolving rapidly to meet these needs by tailoring training to the abilities of the learner. Adaptive models infer an individual’s current skill level and intelligently select training materials to help them progress to more advanced stages. The models are informed by behavioral data, that is, by performance on a training task. They do not, however, take into account the learner’s neurophysiological state. Especially in such a mentally demanding domain as ATC, neurophysiological biomarkers corresponding with cognitive and affective states are needed to provide crucial context, enabling deeper insights that generate recommendations for more optimal training policies. To address these challenges, Aptima, Inc. the University of Chicago, and Reflexion propose NeuroAdapt, a brain-computer interface that incorporates neurophysiological signals into an adaptive learning framework, enhancing motor and cognitive skills for elite skill training. NeuroAdapt fuses training performance data with rich neurophysiological data to derive deep insights into the learner’s cognitive state. Aptima’s ML-backed adaptive learning algorithms then use this information to control training parameters in real time, tailoring the paradigm to the unique needs of the individual learner. We have devised a theory-based, data-driven approach to measure EEG, EKG, and other neurophysiological signals while the trainee is actively using the