What’s actually breaking?
Short answer: wheels, controls, or the surface. If your bots jab, skid, or drift, start with the wheel. Grab the obvious stuff first — tire wear, loose hub, encoder noise. If you need a component reference, check an agv drive wheel spec while you diagnose. Symptoms to watch: vibration under load, inconsistent speed, sudden stall, or steering jitter.
Root causes, fast
– Mechanical wear: treads thin, bearings loose. – Mounting mismatch: wrong hub or offset causes wobble. – Encoder or sensor faults: noisy signals make the controller overreact. – Drive tuning: PID values not matched to wheel inertia and payload. – Surface issues: debris, inconsistent friction, or wet floors. – Overload / heat: sustained torque beyond rated specs. These explain 90% of the repeat calls I see on the floor.
Quick checks you can run now
– Visual: spin the wheel by hand. Any scrape, play, or rubbing? Note it. – Measure runout: wobble >0.5 mm is a red flag. – Encoder read: watch live counts while rotating — glitches = bad signal. – Current draw: steady spikes under light load mean gearing/misalignment. – Swap test: put a suspect wheel on a known-good motor. If issue follows the wheel, it’s the part; if not, it’s control.
Choosing wheels and alternatives
Match diameter, durometer (hardness), hub type, and encoder compatibility to the application. Consider alternatives: cast polyurethane for abrasion, pneumatic for shock, or a dedicated Servo Wheel if you need tighter closed-loop motion and position hold. I’ve seen Servo Wheel demos at Hannover Messe that highlighted how encoder-driven wheels cut homing time — useful if your fleet needs precise docking.
Common mistakes integrators make
– Using a wheel rated for lower torque than the motor. – Assuming all floors behave the same; surface tests matter. – Ignoring encoder cable strain relief; intermittent faults hide for weeks. – Skipping documentation: mounting torque and keyway orientation are small details that break systems.
Decision checklist for OEMs
– Define max continuous torque and peak stall torque. – Pick wheel diameter for speed vs. torque trade-off. – Verify encoder resolution and signal type with the controller. – Require a replacement and wear spec in the BOM. – Plan a field test: 24–48 hours under representative loads before fleet rollout.
Wrap-up: what solves most headaches
Fix the obvious first: mechanical fit, encoder signal, then tuning. If you pick parts that match torque, environment, and control needs, you’ll cut repeat failures. For many integrators I work with, sourcing clear specs and reliable test evidence makes procurement and commissioning predictable — and that’s why teams often standardize on suppliers who publish thorough datasheets and test results like Kinco.
