Robotics ML Engineer
Robotics and ML engineering internship at Robogio Dynamics, building industrial automation systems deployed at facilities of 100-500 employees and discovering a novel emergent phenomenon in ML filter space that increased ARC-AGI 2 accuracy and was flagged by the CEO for integration into Robogio's automated robotic machines. Work spanned low-level firmware on Teensy 4.1 microcontrollers, FOC motor control, three-motor delta actuator systems, and MDL-based vision models.
Key Events Led
MDL Pattern Discovery: ARC-AGI 2
Increased ARC-AGI 2 accuracy; CEO-flagged for production deployment
Identified 'Growing Neighbors', a never-before-described filter activation cascade in MDL space at the intersection of information theory, computer vision, and abstract reasoning, attracting university research interest for potential publication
View project →
Full-Scale Industrial Sanding Robot
Purchased and deployed at a 100-500 person facility; validated by aerospace engineers
Built a complete multi-axis FOC motor control system from scratch with a real-time Web Serial dashboard, automating a previously manual industrial process now active across a professional aerospace supply chain
View project →
Autonomous Multi-Axis Force Compliance System
Self-calibrating delta actuator sourced for Techno Coatings facility-wide deployment
Engineered a three-motor delta actuator with 3x Teensy 4.1s, 3x SimpleFOC shields, and 3x absolute encoders — stall-detection homing and a PD force-compliance controller that actively yields under load, delivering consistent pressure on uneven rotating surfaces and sanding objects without fastening or adjustment
View project →
Key Achievements
- Industrial sanding robot purchased and actively deployed at a facility of 100-500 employees, validated by Techno Coatings aerospace mechanical and electrical engineers, directly automating a previously manual process across a professional aerospace supply chain
- Discovered "Growing Neighbors", a novel emergent phenomenon in MDL filter space that increased ARC-AGI 2 accuracy and was probed by the CEO for real-time deployment across Robogio's automated robotic machines, with a university professor in Upstate New York pursuing potential research collaboration and publication
- Engineered a three-motor delta actuator (3x Teensy 4.1, 3x SimpleFOC shields, 3x absolute encoders) with stall-detection homing and a PD force-compliance controller capable of sanding any object placed on it without fastening or adjustment, work typically performed at undergraduate and graduate level, evaluated by Techno Coatings and sourced for facility-wide deployment
