PADRE
Predicting Actions and Detecting Real-Time Errors
As automation, robotics, and AI transform the job market, companies face an urgent need to reskill and upskill employees—especially in technical and high-risk domains where manual precision and cognitive readiness are critical. Traditional one-on-one training methods are increasingly difficult to scale and often rely heavily on scarce experienced workers.
The PADRE project addresses this challenge by developing a sensor- and AI-driven framework for monitoring, evaluating, and supporting skill acquisition in real time. By capturing fine-grained actions, behavior, and cognitive load during complex tasks, PADRE aims to assess performance quality, track learning progress, and predict potential human error in real-world conditions.
PADRE explores, develops and integrates novel sensing technologies, including workplace-approved glove-based and ultrasonic forearm motion tracking for fine-grained hand and manual action capture, as well as in-ear biosignal monitoring (EEG, EOG, PPG) to assess cognitive load, focus, and hesitation.
The project focuses on high-stakes environments such as industrial chemistry, semiconductor production, pharmaceutics, and advanced medical fields. Beyond specific use cases, PADRE will develop reusable foundation models that can support future training, on-the-job guidance, and safety-critical decision-making across sectors.