Embodied computation
Which decisions can be made by mechanism shape, spring behavior, or contact geometry?
Research agenda
Intelligence you can touch.
How much useful behavior can be designed into a machine’s geometry, materials, and mechanisms before software has to intervene?
Mechanical intelligence
Kyle’s central research interest is mechanical intelligence: the use of physical form, compliance, underactuation, passive dynamics, cams, linkages, and material behavior to perform functions that might otherwise demand additional sensing, computation, or control complexity.
The aim is not to reject electronics. It is to assign each part of a system the kind of work it performs best—creating robots that are more affordable, understandable, repairable, and resilient.
Read the research directionLines of inquiry
Questions worth building.
The work moves between historical mechanisms, modern robotics, rapid prototypes, and measurable performance.
Accessible humanoids
Can underactuation and modular construction make expressive humanoid research platforms substantially cheaper?
Historical machines
What can medieval and Renaissance automata teach contemporary roboticists about compact mechanical programming?
Publication status: Research pages describe active directions and works in progress. Formal papers, datasets, and citations will be linked here when released.