AMRDriveUnitCompare six AMR/AGV layouts. Estimate wheel-output torque, gear ratio and traction DC current, then copy a concept brief for engineering review.
Concept estimates with illustrative assumptions. Safety, traction and component ratings require vehicle-specific validation.
All fields are required and prefilled. Expand assumptions to review every default. Payload is carried mass; trailers need a separate drawbar model.
Rated payload carried by the vehicle frame. Range: 0–30000 kg.
Structure, battery pack, sensors, and passive wheels. Range: 30–8000 kg.
Governed maximum linear velocity on level floor. Range: 0.2–4 m/s.
Two independent fixed traction drivetrains with passive spring-loaded casters. Simple, cost-effective, zero-radius spin capability.
Required for nonzero radius; must not exceed max travel speed. Range: 0.05–4 m/s.
Estimate above floor; confirm with the worst payload position. Range: 50–2500 mm.
Front-to-rear support spacing for the load-transfer screen. Range: 300–4000 mm.
Used only for differential drive at zero turning radius. Range: 0.05–2 rad/s.
Ramp-up duration from standstill to maximum speed. Range: 0.3–8 s.
Steepest operational slope (100 × rise / run). Range: 0–20 %.
Outer diameter of the primary traction wheels. Range: 100–450 mm.
Radius to vehicle center (0 = zero-radius pivot/spin on spot). Range: 0–12 m.
Lateral centerline spacing between left and right drive wheels. Range: 400–2500 mm.
Continuous rated speed ceiling before field-weakening. Range: 30–6000 RPM.
Operating temperature in warehouse or cold store. Range: -40–55 °C.
Recorded in the brief; thermal duty cycle must be measured separately. Range: 4–24 h/day.
Equal torque sharing across 2 driven wheels, 1850 kg gross mass. Outputs describe demand, not an approved motor or vehicle.
Wheel-output torque / drive
39.7 Nm
Steady uphill; peak 178.2 Nm
Concept gear ratio
20.94:1
143 wheel RPM; 3000 motor RPM
Traction peak DC current
134.1 A
6.44 kW at 48 V nominal; auxiliaries excluded
Traction steady DC current
29.9 A
1.14 kW useful steady mechanical power
Static normal load / drive
6347 N
Assumed 70% driven weight share
Minimum wheel load screen
N/A — geometry needed
Only centered, equal four-corner supports, 100% driven, on level floor
Example network target
1000 µs
Example jitter ≤ 20 µs; within assumed budget. Measure actual response.
Translational kinetic energy
2081 J
Excludes rotation, descent and brake duration; not a resistor rating
No example traction threshold exceeded. Grip, stability and dynamic loads remain unverified.
Required friction screen: 0.28 vs assumed dry grip 0.65. Includes lateral demand; layouts outside the four-corner model use static loads and can understate demand. Zero-load results mean contact is lost.
Differential outer / inner turn speeds: 0.64 / 0.36 m/s. Ratio uses the largest straight/turn wheel speed. Pivot yaw is entered separately; pivot slip and rotational energy are not modeled.
Tread scenario assumptions: rolling coefficient 0.012, speed screen 3.5 m/s, cold screen -40°C. No threshold exceeded; these are not supplier ratings.
Model efficiencies: gearbox 0.94, motor 0.92, inverter 0.96, layout factor 0.96, rotating inertia allowance 15%. No measured thermal duty, auxiliary power, battery sag or turning scrub is included.
This tool does not determine stopping distance, scanner fields, suspension stroke, bearing life, SELV/PELV class or CE conformity. Next: send wheel coordinates, CG range, duty trace and battery limits with this brief for review.
Published by AMRDriveUnit. Page created , reviewed . Scope: custom drivetrains within one AMR/AGV vehicle, rather than individual wheel design or fleet traffic control. Presets and worked examples are hypothetical, not customer deployments.
Evidence / scope: ROS Controls — Differential drive controller implementationEtherCAT Technology Group — Distributed Clocksmaxon — What is the meaning of STO?
Two independent fixed traction drivetrains with passive spring-loaded casters. Simple, cost-effective, zero-radius spin capability.
Two steer-drive units can reorient for lateral travel. Wheel locations, steering range, and caster geometry determine achievable maneuvers.
Four corner steer-drive units enable flexible travel directions after steering alignment, with additional controls and commissioning work.
Four fixed drives with angled free rollers permit lateral travel. The model uses a provisional 0.70 aggregate factor, not a universal loss from roller angle.
Single centralized steerable drive unit in front or rear with two passive fixed/caster load wheels. Classic tugger/tow tractor layout.
Six driven wheels with articulated walking beams can improve contact on uneven floors. Load sharing and skid-turn losses require a geometry-specific model.
Schematics identify active wheel count and steering capability. Passive supports, suspension and installation coordinates are omitted.
| Layout | Useful when | Cost / complexity to budget | Main limitation to test |
|---|---|---|---|
| Differential 2WD | Forward travel and chassis pivoting | Two traction axes plus passive supports | No lateral translation; caster behavior and drive load share |
| Dual steer-drive | Lateral alignment after steering reorientation | Two traction and two steering axes | Steering sweep, caster constraints and transient scrub |
| Quad steer-drive | Independent wheel alignment and versatile routes | Four traction and four steering axes | Calibration, steering transitions and load sharing |
| Mecanum 4WD | Lateral travel without steering actuators | Four traction axes and roller maintenance | Floor joints, roller contact, directional losses and vibration |
| Tricycle | Simple forward/reverse routes with swept turns | One traction and one steering axis | Swept path, drive-wheel loading; center pivot not assumed |
| Rocker-bogie 6WD | Contact retention on uneven routes | Six traction axes plus articulation | Actual linkage loads, turning scrub and joint durability |
These are candidate directions. Choose between them with a swept-path simulation and loaded floor trial; the tool does not rank architectures by a fabricated score.
| Output | Equation | Interpretation / limit |
|---|---|---|
| Gross mass m | payload + chassis (kg) | Carried payload only; trailer drag is not supported |
| Steady force F | m × g × (Crr × cos θ + sin θ); θ = atan(grade/100) | Straight uphill rolling and grade resistance |
| Peak force Fp | F + 1.15 × m × (vmax / acceleration time) | 15% rotating-inertia allowance; not a measured acceleration trace |
| Wheel torque T | F × wheel radius / (driven wheels × layout factor) | Wheel-output demand, not motor-shaft torque; assumes equal sharing |
| Wheel speed n | 60 × wheel surface speed / (π × diameter) | Differential uses greatest straight/turn wheel speed; other layouts straight-line only |
| Ratio i | motor rated RPM / required wheel RPM; direct drive = 1 | Concept ratio needs an available reducer and motor torque-speed curve |
| Electrical power | F × vmax / (layout × gearbox × 0.92 × 0.96) | Motor/inverter efficiencies are fixed examples, not low-speed performance maps |
| DC current | electrical watts / nominal DC voltage | Traction-only bus demand; phase current and battery sag excluded |
For a wheel-output torque T and reducer ratio i, the motor torque candidate is T / (i × gearbox efficiency). Match its continuous and transient torque-speed envelope to the actual duty. Do not buy a motor from peak watts alone. Worst grade, maximum speed and acceleration are combined here as a demand envelope; verify whether the application requires them simultaneously.
For a two-wheel differential base, outer/inner wheel speeds are v ± ω × track/2. On a circular path, ω = v/R. If R is less than half the track, the inner wheel reverses. At R = 0 the center is stationary: the entered pivot yaw rate supplies ω rather than reusing maximum travel speed.
Evidence / scope: ROS Controls — Differential drive controller implementation
The largest wheel speed sets the differential speed/ratio envelope. A short radius at nonzero center speed can demand an impractically high outer-wheel speed; reduce the entered turn speed and verify the motor ceiling. Roller heat, caster drag and pivot torque are not included.
Steer-drive units require wheel coordinates, steering limits and orientation. Mecanum requires roller orientation and wheelbase geometry. Bogie vehicles need skid-turn resistance. Their turning wheel speeds are shown as N/A, with straight-line RPM retained for preliminary sizing. A conventional tricycle must use a positive turning radius; its actual minimum radius still needs steering geometry.
| Quantity | Method | What remains unknown |
|---|---|---|
| Static drive-wheel normal load | m × g × cos θ × driven share / drive count | Actual static CG position and support stiffness |
| Front/rear transfer | m × longitudinal acceleration × CG height / wheelbase | Allocation to specific wheels and grade-induced moment |
| Inside/outside transfer | m × (turn speed² / R) × CG height / track | Roll compliance, transient response and off-center payload |
| Minimum wheel load screen | m × g / 4 − longitudinal transfer/2 − lateral transfer/2 | Only equal four-corner, centered CG, 100% driven, level floor |
| Friction demand screen | hypot(longitudinal wheel force, lateral wheel force) / normal load | Measured friction circle, unequal torque and floor contamination |
The four-corner screen combines the maximum straight acceleration and entered turn acceleration as a stress envelope. It is not an exact turn trajectory. Zero load means loss of contact; below the model assumptions, no wheel grip approval is possible. For other support layouts the friction screen uses static load and can understate demand, so obtain wheel-specific loads before specifying traction.
Lower CG, wider supports or reduced corner speed can reduce load transfer, but suspension changes must be validated with the payload range and actual floor. Do not add suspension merely to compensate for an unknown CG position.
| Protocol option | Example cycle / jitter | Required evidence before selection |
|---|---|---|
| EtherCAT DC | 250 µs / 0.1 µs | Drive DC support, controller schedule, timestamped following error |
| CANopen | 1,000 µs / 15 µs | Bitrate, PDO mapping, node count, SYNC behavior and bus-load trace |
| PROFINET IRT | 250 µs / 0.5 µs | IRT-capable devices, configured schedule and application response |
| Modbus TCP | 10,000 µs / 2,000 µs | Supervisory use only in this screen; no deterministic multi-axis assumption |
The example topology budgets are 1,000/20 µs (cycle/jitter) for differential, tricycle and bogie; 500/5 µs for dual steer-drive and Mecanum; 250/1 µs for quad steer-drive. “Within assumed budget” only compares these two sets of example values. Controller tuning, following-error limits and measured end-to-end latency establish the real requirement.
CiA 402 CSP, CSV and CST modes are candidates for steering position, traction speed and traction torque. Confirm implementation in the selected drive. CST does not provide a ready-made slip controller, and a regular motion bus does not confer a safety integrity level.
Evidence / scope: EtherCAT Technology Group — Distributed ClocksLenze i950 — CiA 402 operating modes
| Choice | Potential benefit | Evidence to request / cost to allow |
|---|---|---|
| Coaxial planetary | Conventional reduction in a narrow radial envelope | Continuous/peak torque, radial load, ratio availability and efficiency map |
| Bevel-planetary | Can move motor length out of the vertical envelope | Full installation drawing, bevel stage losses, mounting and service access |
| Cycloidal | Candidate where stiffness and shock loads matter | Actual backlash, overload duty, mass and lubrication; no universal 500% rating |
| Direct-drive hub | No gear reduction or gear backlash | Low-speed torque, motor current, heat rejection and bearing design |
| Polyurethane / rubber | Different rolling, grip and noise behavior | Exact load-speed-temperature rating, floor tests and replacement cost |
| Articulation / preload | Can improve wheel contact on an uneven route | Joint survey, spring curves and payload travel; floor standards do not specify stroke |
The tread choices use illustrative rolling coefficients 0.012–0.028 and assumed dry grip 0.60–0.80. These are not measured values for any branded compound. The displayed speed and cold thresholds are example screens, not continuous ratings or glass-transition temperatures. Validate the assembled wheel at actual load and speed; hardness and material family alone are insufficient.
Evidence / scope: Blickle — Polyurethane wheel selection
At equal demanded power, raising nominal DC voltage lowers calculated DC current. It does not establish touch safety or battery suitability. Review maximum charge voltage, bus ripple, transients, protective separation, grounding and isolation. SELV/PELV cannot be assigned solely from nominal voltage.
Evidence / scope: Schneider Electric — SELV circuit separation guidance
Need to review battery and braking limits with the drive selection? Request a drivetrain engineering review
The energy card shows ½ × gross mass × maximum travel speed² in joules. It excludes rotating inertia and downhill potential energy. A resistor needs energy, pulse duration, peak power, resistance, bus voltage and repeat duty; this energy value alone is insufficient. Scheduled hours are recorded in the brief but do not establish motor temperature or bearing life.
| Risk | Decision impact | Mitigation / alternative |
|---|---|---|
| Misuse: static traction treated as dynamic approval | Wheel unloading or slip despite sufficient torque | Measure per-wheel loads and floor grip; reduce speed until validated |
| Cost: too many active axes | Extra drives, cables, spares and commissioning time | Compare differential with steering layouts against actual docking routes |
| Scenario mismatch: trailer modeled as carried load | Incorrect drive normal force and drawbar demand | Separate tractor/trailer rolling and hitch-load model |
| Battery rejects regeneration | DC overvoltage or loss of commanded braking | Test high SOC/cold battery behavior; design a validated absorption path |
| STO mistaken for a brake | Coasting or rolling after torque removal | Validate coordinated stopping and independent parking/holding behavior |
| Network or encoder fault | Steering/traction disagreement and tire scrub | Measure following error and test fault response with selected devices |
| Floor joints or compound mismatch | Wheel shock, loss of contact and replacement expense | Loaded route test; compare preload, compliant wheels and route speed limits |
ISO 3691-4 addresses vehicle safety requirements and verification. The public standard abstract confirms that scope; this page does not interpret unverified clauses. Select safety functions and validate the whole system using the applicable standards and component documentation.
STO removes torque generation; it does not establish a defined stop, ramp holding, electrical isolation, or a safety certification. Commission braking and parking as separate behaviors, including power loss and loss of charge acceptance.
Evidence / scope: ISO 3691-4:2023 — Driverless industrial trucks and their systemsmaxon — What is the meaning of STO?
As reviewed on 5 November 2026, Regulation (EU) 2023/1230 generally applies from 20 January 2027. A drive module is not automatically partly completed machinery. The delivered function and applicable legislation determine classification and documentation; have the conformity assessment reviewed for the actual product.
Evidence / scope: EUR-Lex — Machinery safety requirements and transition dates
These examples use the named calculator presets and the method above. They are not comparable bids: masses, grades and wheel sizes differ. The purpose is to prepare engineering questions, not to rank designs by a single number.
Premise: 1850 kg gross, 1.5 m/s, 1.2 s acceleration, 3% grade; 200 mm wheels and 2 drives.
Process: calculate rolling + grade + inertia force, split wheel torque using the layout factor, then divide power by the assumed stage efficiencies.
Concept output: 178.2 Nm peak per drive; 20.94:1 ratio; 134.1 A peak traction DC.
Next test: Confirm caster-supported drive loading and outer-wheel speed on the actual route.
Premise: 4350 kg gross, 1.2 m/s, 1.8 s acceleration, 5% grade; 250 mm wheels and 2 drives.
Process: calculate rolling + grade + inertia force, split wheel torque using the layout factor, then divide power by the assumed stage efficiencies.
Concept output: 397.4 Nm peak per drive; 32.72:1 ratio; 197.8 A peak traction DC.
Next test: Provide wheel coordinates and steering limits; check docking transitions and scrub.
Premise: 1050 kg gross, 1 m/s, 0.8 s acceleration, 0% grade; 150 mm wheels and 4 drives.
Process: calculate rolling + grade + inertia force, split wheel torque using the layout factor, then divide power by the assumed stage efficiencies.
Concept output: 47.0 Nm peak per drive; 1.00:1 ratio; 61.0 A peak traction DC.
Next test: Validate roller vibration, floor joints and hub thermal duty. No cleanroom suitability is asserted.
Compare steer-drive when the route needs lateral alignment without rotating the chassis. There is no universal payload or aisle-width threshold. Differential drive uses fewer active axes; steer-drive adds steering clearance, alignment time, and commissioning work.
No. Conventional steer-drive wheels usually need to align before a new travel direction. Steering limits, wheel locations, and transition time must be included in the route simulation. Four traction and four steering axes also add hardware and calibration work.
No. It is a provisional aggregate scenario factor, not a universal 30% loss from roller geometry. Measure direction-dependent force, roller losses, and floor-joint behavior for the chosen wheel. The calculator does not solve Mecanum inverse kinematics.
This tool assumes carried payload. A trailer changes hitch force, rolling losses and grade demand without adding equivalent normal force to the drive wheels. Use a separate drawbar and hitch-load model; the tricycle example represents carried gross mass only.
Wheel speed is v plus or minus yaw rate times half the track. When the center turning radius is less than half the track, the inner wheel reverses. At zero radius, the entered pivot yaw rate determines equal and opposite wheel speeds.
No. They are example screening budgets. EtherCAT Distributed Clocks can synchronize device clocks precisely, but controller scheduling, drive response, network load and application error limits determine the actual cycle budget. Measure the configured network before release.
CSP is a candidate for position-controlled steering, CSV for traction speed and CST for torque control. Verify supported modes in the drive manual. Selecting CST alone does not implement slip control; that needs feedback and a validated control strategy.
No. The result is traction input power divided by nominal DC voltage. Motor phase current needs torque constants and control data. Battery sag, steering, lifts, computers, cooling, charge transients and auxiliary loads are excluded.
No. Protective separation, grounding, maximum operating voltage, transients and the applicable environment matter. SELV and PELV are circuit protection concepts, not categories assigned simply by splitting nominal voltages at 60 V.
Per-wheel allocation is only screened for a centered, equally supported, fully driven four-corner layout on level ground. Caster geometry, articulated suspensions, noncentral CG or grades require a geometry-specific model. N/A means more data is needed, not that traction is acceptable.
No. Survey joints, local steps and waviness along the wheel paths, then combine wheel spacing, tire compliance, spring preload and payload travel. A building flatness designation alone does not determine a suspension stroke or prove contact retention.
No. Thresholds are illustrative screening assumptions. Request the exact wheel load-speed-temperature envelope and test wet, dusty or cold floors. No warning does not establish thermal durability, a glass-transition temperature, or adequate traction.
Hours are recorded for the engineering brief. Continuous torque approval needs an actual speed/torque duty trace, RMS heating and cooling data. This tool does not calculate bearing life or infer motor temperature from daily hours.
No. ISO 3691-4 covers whole-vehicle safety verification. Measure worst-case braking, sensor and controller response, slope behavior and traction, then validate the protective-field design. No standards-compliant distance is calculated here.
STO prevents motor torque generation; it does not define a controlled stop or holding behavior. Validate coordinated stopping and parking separately, including battery charge rejection and loss of power. An unchecked holding-brake option is not a safety approval.
No automatic classification is made here. Assess the delivered product, independent function, intended incorporation and applicable legislation. As reviewed on 5 November 2026, the Machinery Regulation generally applies from 20 January 2027. Obtain product-specific conformity advice and documentation.
Source review date: 2026-11-05. Primary references support only their stated scope. Manufacturer examples do not establish ratings for other products. Calculator coefficients, timing budgets and scenarios are disclosed assumptions. No public evidence substantiates customer wear reduction, uptime, stopping performance or cleanroom certification for the examples here.
AMRDriveUnit maintains this method. Review every six months, or earlier when component documentation, calculation assumptions or applicable standards change. Page creation records repository provenance; it is not a verified public publication date.
Public abstract confirms safety requirements and verification scope for driverless industrial trucks. Full clauses are not reproduced or used to certify calculator outputs.
Primary controller implementation for signed left/right wheel speeds from body linear and angular velocity. It does not validate traction or vehicle stability.
Describes distributed clock synchronization below one microsecond. Device sampling synchronization is distinct from end-to-end loop response; no universal topology-specific cycle requirement is established.
Manufacturer documentation lists cyclic synchronous position, velocity and torque modes. Actual support and configuration depend on the selected drive.
Explains why torque removal does not define deceleration, holding capability, or electrical isolation. Device certification and a separate braking concept may be required.
Explains protective separation from higher-voltage circuits and separation from earth for SELV. It does not classify a complete mobile robot circuit from nominal battery voltage or prescribe a PELV design.
Manufacturer overview of polyurethane tread families and selection properties. It does not substantiate the calculator example grip coefficients, temperature thresholds, or continuous speed ratings.
Official summary states general application of Regulation (EU) 2023/1230 from 20 January 2027. Product classification, applicable legislation and documentation require a product-specific assessment.
Copy the concept brief above, add wheel coordinates and loaded CG range, route and floor measurements, speed/torque duty trace, battery/BMS limits and stopping requirements. Paste it into the inquiry form message.