Drivetrain 'marriage': engine-to-transmission/axle spline mating and torqued joining
A signature step in large-tractor assembly is the 'marriage' of the engine (from Engine Works) to the transmission and axle assemblies, where heavy modules are lowered and bolted into an integrated drivetrain. The contact-rich core of the task is blind spline/coupling alignment — engaging input shaft splines and locating dowels before the housings can mate flush — followed by sequenced, torqued fastening. The components are large and heavy and the mating interface is largely occluded once parts are close. It is hard for a robot because final spline engagement is felt (slight rotation-plus-push until teeth mesh) rather than seen, and forcing misaligned splines damages gear teeth or bearings. Payload here far exceeds typical dexterous-arm range, making this a borderline AGD fit (the manipulation intelligence is relevant even if the lift is not), and no public signal indicates intent to automate it — demand is pure inference. We identified this through our own research; we have not confirmed the specifics with the customer directly. This page is our researched read — a starting point for that conversation.
What the task is
RESEARCHED · our reconstructionA signature step in large-tractor assembly is the 'marriage' of the engine (from Engine Works) to the transmission and axle assemblies, where heavy modules are lowered and bolted into an integrated drivetrain. The contact-rich core of the task is blind spline/coupling alignment — engaging input shaft splines and locating dowels before the housings can mate flush — followed by sequenced, torqued fastening. The components are large and heavy and the mating interface is largely occluded once parts are close. It is hard for a robot because final spline engagement is felt (slight rotation-plus-push until teeth mesh) rather than seen, and forcing misaligned splines damages gear teeth or bearings. Payload here far exceeds typical dexterous-arm range, making this a borderline AGD fit (the manipulation intelligence is relevant even if the lift is not), and no public signal indicates intent to automate it — demand is pure inference.
Is this the actual task and sequence? What are the real tolerances, cycle rate, and reject criteria, and which steps are today's manual bottleneck? Answering these is what turns this from a researched signal into a validated use case.