AGV Drive Wheel Gearbox Supplier: Sizing & Qualification

1. What to decide before shortlisting a supplier

This page is a buyer screening method, not a supplier ranking. Its calculations are transparent first-principles estimates; the evidence below supports the review method, not a promise that any particular gearbox meets it.

2. Reproduce the tool results and assumptions

Units: mass m in kg; g = 9.80665 m/s²; diameter D in mm; speed v in m/s; acceleration a in m/s²; n driven wheels; η as a fraction. Grade and forward acceleration are assumed simultaneous. Rolling coefficient 0.02 and efficiency 94% are editable example assumptions, not material or product ratings.

Exclusions: no motor/rotor inertia, aerodynamic resistance, trailer drag, turning, floor-joint impact prediction, downhill stopping or thermal time model. A displayed acceleration demand is not a gearbox peak capacity, and the steady uphill value is not an RMS duty rating.

3. Compare drive architectures at the same operating points

The following is an engineering review checklist, not a performance ranking. No universal vehicle-mass cutoff establishes whether a direct drive or reducer is suitable. Ask each bidder for its configured torque-speed-temperature envelope.

Catalog performance, prices, acoustic levels and lead times are configuration-specific and are unknown here. Obtain equivalent dated offers instead of assuming that one architecture is always lighter, cheaper or thermally superior.

4. Wheel loads, bearing life and preload evidence

The tool estimates Fstatic = m g × driven weight share × cosθ / n; Fdynamic = Fstatic × entered multiplier; M = Fdynamic × L. These are external reactions at a stated plane. Internal bearing forces depend on support spacing, contact angles, axial loads and load distribution. A dedicated wheel-drive product illustrates why the output-bearing design deserves separate review.

SKF explains L10 = (C/P)ᵖ in millions of revolutions and L10h = 10⁶ L10 / (60 N), with p = 3 for ball bearings and 10/3 for rollers. Use bearing speed N, not automatically motor speed. The tool lacks bearing geometry, C and equivalent bearing load P, so it deliberately does not display a life estimate.

SKF recommends calculation and application testing for new preload designs. A generic negative clearance or cold starting torque window cannot qualify every hub size. ISO 281 addresses bearing fatigue; it is not a guarantee against wear, leakage or corrosion.

5. Backlash and turning: build a vehicle error budget

Navigation type is RFQ context only. Allocate permissible wheel motion error from docking accuracy, encoder placement, controller behavior, tire compliance and reversal frequency. A LiDAR or QR label alone does not establish a gearbox backlash limit.

For illustration, a 200 mm wheel with 6 arcmin output angular deadband corresponds to rθ = 100 mm × (6π / 10,800) ≈ 0.175 mm of rim displacement for that angle. This is not a prediction of vehicle localization drift; unequal wheel errors, floor slip and sensor fusion change the outcome.

Turning adds tire-floor shear and steering-axis moments. The previous single contact-patch expression cannot be applied directly as wheel-shaft torque without track width, contact geometry and a vehicle force model. Request a measured turning-current/torque trace at maximum laden mass and validate the gearbox against repeated reversals.

6. Qualify thermal duty, seals and lubricant together

Supply a torque-speed-time cycle, hours per shift, enclosure temperature, wheel loads and mounting heat path. Require a loaded run at representative conditions, including cold start where applicable. A motor duty label or high catalog efficiency alone does not establish gearbox thermal capacity.

7. Braking and stopping qualification

ISO 3691-4:2023 covers safety requirements and verification for driverless trucks and their systems. Its public scope is not evidence for a fixed 2.8× gearbox torque requirement. This page does not calculate a stopping torque or claim compliance.

Send the supplier the worst downhill/loaded condition, speed, required deceleration, stopping distance, brake location, reflected inertia, drive control behavior and expected stop count. Capture transient torque and tire slip during vehicle tests. A motor-mounted brake, regenerative stop and mechanical wheel brake can impose different drivetrain loads.

Specify both normal and emergency stopping tests with the vehicle safety engineer, using the applicable full standards and project requirements. Turning and braking acceptance remain open until those tests pass.

8. A supplier scorecard based on returned evidence

No published failure dataset, customer warranty audit or verified competitor quote supports a numerical supplier ranking here. Mark missing evidence as not demonstrated; do not award an assumed pass.

Robotized Gearbox is a candidate for engineering review, not an automatically qualified match. Ask for a configured proposal and evidence against this same matrix. The Neugart reference is an example of public product documentation, not an assertion of equivalence.

9. Compare lifecycle cost with explicit scenarios

Hypothetical buyer arithmetic in USD, not market pricing or field reliability data. Assume a five-year / 20,000-hour horizon. Replacement counts below are sensitivity inputs; no supplier is assigned a failure rate. Taxes, discounting, energy and residual value are excluded.

The lower purchase price wins if neither fails under these assumptions; different replacement outcomes can reverse the decision. Request field evidence before assigning probabilities. Include tooling, spare inventory, freight and maintenance in your actual purchasing model.

10. Worked screening scenarios, not customer case histories

11. Send a complete supplier RFQ

Use the tool’s copy/download output as a preliminary attachment. Unknown loads, braking, preload, life and product capacity must stay marked as unresolved until evidence is returned.

12. Sources, review date and evidence limits

Public primary sources checked on November 1, 2026. Standards editions and product data can change; confirm applicability at design release. Public abstracts do not replace the full standards. No independent gearbox test dataset or comparative price survey is available for this page.

1. 1. ISO 281:2007 — Dynamic load ratings and rating life

   Basic and modified bearing rating life; basic life uses 90% reliability. Does not establish a complete wheel-hub load model or guarantee service life.

2. 2. ISO 3691-4:2023 — Driverless industrial trucks and their systems

   Vehicle/system safety requirements and verification. The public scope does not establish a universal gearbox emergency-stop torque multiplier. Confirm the applicable edition for your project.

3. 3. SKF — Bearing preload (technical guidance, PDF)

   Preload depends on design, operating temperature and housing elasticity. New designs require calculation and application testing; no universal micrometre or drag-torque window is adopted here.

4. 4. SKF — Basic rating life (catalogue excerpt, PDF)

   L10 and L10h equations, ball/roller exponents and the influence of lubrication, contamination and installation on actual life.

5. 5. Neugart — NGV application-specific gearbox

   A named manufacturer example of a wheel-drive gearbox with a dedicated output bearing arrangement. Ratings and load reference geometry must be read for the configured size; this is not a rating for Robotized Gearbox products.

The architecture matrix, scorecard and cost scenarios are editorial decision aids. Numerical defaults are transparent assumptions; the worked examples are reproducible calculations. They are not field observations, supplier guarantees or proof of product compliance.

13. Supplier qualification FAQ

14. Next step: turn the screen into a qualified proposal

Attach your preliminary RFQ, drawings and missing-data list. Ask each supplier to return the configured performance envelope and a sample test plan before ordering production units.