UGV Autonomy Engineer

Penn’s end-to-end robotics intro — perception, planning, mechanics.

Waymo-grade perception patterns port straight to UGVs.

Nav2 is the default autonomy stack for most UGV research + startups.

SLAM + particle filters are even more critical on the ground.

Off-road, "is this path drivable?" is the whole planning problem.

Non-convex, high-dim terrain planning is where sampling planners shine.

Realistic off-road physics at scale — hard to get anywhere else.

Gets a lot of UGV autonomy code wrong when people forget this.

Before you drive, you want a bench that proves the full loop works.

Ground-to-ground + ground-to-air teaming lives on tactical mesh.

Interop standard — NATO customers will require it.

Ground robots operate in jammed RF and degraded GPS environments more often than air platforms. Dead-reckoning, visual-inertial nav, and graceful comms-loss behavior are the field-realistic UGV problems.

If you run UGV test campaigns, OSHA-30 shortens a lot of PPE conversations.