Gennaro Scarati

• Develop and deploy AI and robotics systems using ROS2, C++, and Python, integrating behavior trees to manage complex task execution and recovery measures.

• Implement learning-from-demonstration pipelines, including diffusion policies, for dexterous manipulation tasks in industrial and agricultural settings.

• Design and integrate computer vision and AI algorithms for detection, segmentation, and tracking within manipulation and navigation pipelines.

• Lead robotics projects for private clients (startups, multinational companies), and public R&D programs, managing systems requirement definition and technical interactions to ensure alignment with projects goals.

• Extensive hands-on experience with collaborative robots (KUKA, Universal Robots, Ufactory, etc.), RGB-D and thermal cameras, diverse end-effectors (vacuum, magnetic, finger grippers), and various mechatronic systems.

• Contractor For General Motors.

• Defined specifications and developed control and diagnostic systems for General Motors vehicles (GMC Sierra and Chevrolet Silverado, 2024–2027), with projected deployment on around 100,000 vehicles in 2024 alone.

• Improved engine diagnostic systems using Six Sigma analysis, machine learning estimation techniques, and genetic programming for calibration optimization.

• Collaborated with cross-functional teams in Italy, the U.S., and India to meet development, calibration, and production milestones.

• Defined control systems and testing requirements in an Agile (Scrum) environment, leveraging the V-Model lifecycle and Model-Based Systems Engineering (MBSE) methodologies.

• Developed automated testing methods for control algorithms and calibrations in simulation environments, complemented by test bench validation and both SIL and HIL testing to ensure robust system performance.

• Developed and validated an UAV Autonomous Landing Technology for precise landings on moving platforms, including the full design, simulation, implementation and experimental testing of localization techniques (UWB and custom Kalman Filters), control algorithms, and state machines to manage all chasing and landing phases.

• Achieved centimeter-level landing precision in real-world field testing with a custom UAV and rover, following extensive simulation using real sensor noise data and subsequent implementation and validation of the developed systems in real scenarios.