A mobile robot is a chain of compromises where the motor, gearbox, controller, battery, wheels, and mechanical layout constantly influence one another. Looking at any of them in isolation almost always leads to a second design iteration.
Faisal Mahmood, Technical Content Writer and CEO | Ourex.pro
Most motor selection mistakes don't reveal themselves during design reviews. They wait until the prototype starts behaving like a real machine instead of a CAD model.
The robot reaches its target speed during the first test, the current draw looks reasonable, and everyone leaves the lab thinking the drivetrain is in good shape. A week later the same robot is carrying its full payload, stopping every few meters, climbing loading ramps, and repeating that cycle for hours. Now the motors run hotter than expected, acceleration becomes inconsistent, and the controller begins reducing output to protect itself. Nothing is technically broken, yet the robot no longer behaves the way the calculations suggested.
That disconnect is common because motor selection rarely fails due to one bad component. More often, it fails because the drivetrain was optimized around a single specification instead of the way the entire machine works. A mobile robot is a chain of compromises where the motor, gearbox, controller, battery, wheels, and mechanical layout constantly influence one another. Looking at any of them in isolation almost always leads to a second design iteration.
Start with the Load, Not the Motor
Engineers naturally enjoy comparing motor specifications, but the motor should almost be the last thing you choose.
The first numbers worth calculating belong to the robot itself. Total operating mass, expected payload, maximum acceleration, allowable stopping distance, wheel diameter, floor conditions, ramp angles, and daily operating hours define the problem far better than any motor catalogue can. Those values establish the forces the drivetrain must produce before electrical components even enter the discussion.
One design review taught this lesson rather well. The prototype comfortably handled its rated payload across a smooth factory floor, yet struggled whenever it approached the loading dock. The motors weren't undersized. The original calculations simply assumed level travel. A five-degree incline increased the required tractive effort enough that the motors spent much longer near their current limit than anyone anticipated. The fix wasn't a larger motor. It was revisiting the entire drivetrain ratio because the operating environment had quietly changed the design requirements.
Think in Wheel Torque Rather Than Motor Power
Motor power is easy to compare. Wheel torque determines whether the robot actually moves.
The drivetrain exists to generate tractive force where the tire meets the floor. Everything between the battery and the wheel simply influences how efficiently that happens. Working backwards from the wheel usually exposes design constraints much earlier than working forward from a motor datasheet.
Start by estimating the force required to accelerate the fully loaded vehicle while overcoming rolling resistance and any expected gradient. Converting that force into wheel torque immediately provides a more realistic design target than selecting a motor based on rated wattage alone. The numbers often surprise people. A drivetrain that appears comfortably oversized at steady speed may have very little margin during repeated starts, tight turns, or incline transitions.
Wheel diameter deserves far more attention than it usually receives because it quietly changes almost every calculation that follows. Increasing the wheel size often begins as a packaging decision. Someone wants better obstacle clearance or smoother travel across uneven flooring. A few weeks later the robot feels noticeably slower leaving a standstill even though the controller tuning hasn't changed. The explanation is mechanical rather than electrical. A larger wheel reduces the available tractive force for the same output torque, which means the entire drivetrain now has less mechanical advantage.
That single decision can force engineers to reconsider gear ratio, controller current limits, and battery capacity long before anyone thinks about replacing the motor.
Continuous Torque Usually Decides Whether the Design Works
Peak torque sells motors. Continuous torque keeps robots operating through an entire shift.
It's tempting to compare maximum torque figures because they provide reassuring safety margins on paper. In practice, most industrial mobile robots spend very little time operating at those limits. They spend hours producing moderate torque while generating heat that has nowhere to disappear quickly.
Prototype testing often exposes this difference. Early performance looks excellent because the windings are still cool and the battery is fully charged. Several hours later the same robot begins limiting acceleration, not because the controller suddenly became inadequate but because thermal conditions have changed. Current remains available. Usable torque does not.
That is why efficiency curves deserve as much attention as maximum output ratings. Two motors capable of delivering similar peak torque can behave very differently during continuous operation. One remains comfortably inside its efficient operating range while the other spends much of the day converting electrical energy into heat.
Bigger Motors Can Create Smaller Margins
Oversizing the motor often feels like the safest decision because additional torque appears to solve future problems before they happen. It rarely works out that neatly.
A larger motor brings higher rotor inertia, increased mass, greater peak current demand, and usually a larger controller. Those changes ripple through the rest of the machine. Battery discharge rates increase during acceleration, cable sizing changes, cooling requirements become more demanding, and available installation space begins disappearing surprisingly quickly.
None of that means larger motors are wrong. Heavy transport robots and outdoor autonomous vehicles frequently need them. The point is that motor size should be the result of drivetrain calculations rather than uncertainty. A properly matched motor operating comfortably within its continuous rating almost always produces a more efficient and predictable machine than an oversized motor spending most of its life well below its intended operating range.
By this point, the discussion has moved well beyond choosing a motor. The required wheel torque is known, realistic operating conditions have replaced optimistic assumptions, and the consequences of wheel diameter, payload, and thermal behaviour are becoming clear. The remaining decisions revolve around how the gearbox, electrical system, and control strategy shape the performance that the motor can actually deliver under real operating conditions rather than ideal laboratory tests.
Gear Reduction Is Part of the Motor Selection
Once the required wheel torque is understood, the temptation is to search for a motor capable of producing it directly. That rarely leads to the most practical drivetrain. In most industrial mobile robots, the gearbox determines how effectively the motor's characteristics are converted into usable wheel torque. A well-matched reduction allows the motor to operate where it is efficient while delivering the low-speed force the application actually needs. A poor ratio does exactly the opposite. The robot reaches its target speed during commissioning but feels strained every time it accelerates, turns with a payload, or climbs a modest incline.
The first prototype often exposes this mismatch long before calculations do. If the motors become warm after repeated stop-and-go cycles while drawing current well below their advertised limits, the gearbox deserves as much scrutiny as the motor. Engineers sometimes assume the motor is undersized when the real issue is that the reduction ratio forces it to spend too much time in an inefficient operating region.
Planetary gearboxes are often chosen for mobile robots because they package high torque capacity into a compact form and distribute load across multiple gears. Spur gearboxes remain a sensible choice where simplicity, cost, and ease of maintenance matter more than maximum torque density. Neither is universally better. The decision depends on the available space, expected shock loading, efficiency targets, backlash tolerance, and maintenance strategy. Looking only at the gearbox ratio without considering mechanical efficiency can create another surprise. Every stage introduces losses, so the wheel never receives all of the motor's theoretical torque.
A Motor Datasheet Is More Than a List of Specifications
The gearbox also changes how the motor should be interpreted. Kv, for example, is often treated as a speed specification, yet it influences much more than maximum RPM. A higher Kv motor generally produces less torque per ampere than a lower Kv design. That doesn't make either one preferable. A high-speed inspection robot with substantial gear reduction may benefit from one approach, while a heavy AMR operating at modest speed may favour another. The important point is to evaluate Kv together with the intended gear ratio rather than as an isolated specification.
The same relationship exists between Kv and Kt. They are linked by the motor's electromagnetic design, which means increasing one inevitably affects the other. During motor selection, Kt becomes particularly useful because it provides insight into how much torque the motor produces for each ampere of current. That information becomes increasingly valuable once controller limits and battery capability enter the conversation.
The Electrical System Determines How Much of the Motor You Can Actually Use
Motor controllers rarely receive enough attention during early drivetrain design. Teams sometimes choose a motor first and assume the controller can simply be sized afterwards. The opposite approach usually produces better results. Continuous current rating, peak current capability, switching frequency, and thermal design all influence how much of the motor's theoretical performance can actually be used. A controller operating near its thermal limit will reduce output long before the motor reaches its advertised capability.
Battery behaviour introduces another constraint that is easy to underestimate. Nominal voltage printed on a datasheet tells only part of the story. Repeated acceleration, cold operating conditions, and aging cells all contribute to voltage sag under load. If the drivetrain depends on maintaining ideal battery voltage to achieve its performance targets, those targets will gradually disappear during normal operation. Designing with realistic voltage conditions rather than laboratory values produces far more predictable results.
Mechanical Decisions Shape Control Performance
Field Oriented Control has become the preferred strategy for many industrial brushless drive systems because it provides smooth torque production across a wide operating range. That does not eliminate the importance of feedback devices. Hall sensors are adequate for many traction applications, but demanding navigation tasks often benefit from higher resolution feedback. The difference becomes noticeable during low-speed positioning, docking manoeuvres, and precise velocity control where encoder quality directly influences control stability.
Incremental encoders are widely used because they balance performance with cost, although they require homing after power cycles in many systems. Absolute encoders remove that requirement and preserve position information immediately after startup. Whether that additional capability is worthwhile depends entirely on the application. Warehouse AMRs operating continuously may justify the investment, while simpler transport platforms may not.
PID tuning often receives attention only after the hardware is assembled, yet mechanical decisions strongly influence how easy the controller is to tune. Gearbox backlash, wheel compliance, drivetrain inertia, and bearing stiffness all appear in the control response. Excessive backlash may look acceptable during manual driving but reveal itself as oscillation during autonomous positioning. Likewise, wheel slip caused by aggressive acceleration cannot be corrected through software alone if the available traction has already been exceeded.
Validate the Drivetrain Before You Trust the Calculations
Endurance testing usually reveals these interactions better than any simulation. A drivetrain that performs perfectly during twenty minutes of laboratory testing may behave differently after eight hours of continuous operation. Bearings warm slightly, lubricant viscosity changes, battery voltage falls, and temperatures rise throughout the enclosure. Small effects accumulate until they influence navigation accuracy, stopping distance, or current consumption. None of these changes are dramatic on their own, but together they determine whether a robot feels refined or constantly in need of adjustment.
Reading a motor datasheet becomes much easier once these relationships are understood. Peak torque attracts attention, but continuous torque deserves greater trust. Maximum current means little without knowing how long it can be sustained. Efficiency curves reveal where the motor is happiest, while winding resistance offers clues about heat generation. Gearbox efficiency deserves the same scrutiny because theoretical output torque is always reduced by mechanical losses before it reaches the wheels.
The engineers who consistently build reliable mobile robots are rarely those who find the most powerful motor. They are usually the ones who keep asking whether every component supports the same operating objective. A slightly different gear ratio may remove the need for a larger battery. Choosing a motor with a more suitable Kv may reduce controller current requirements. Adjusting wheel diameter early in the design may avoid expensive drivetrain changes later.
A practical selection process has served me well over the years. Define the robot's real operating conditions before opening a motor catalogue. Calculate the wheel torque needed under the most demanding realistic load, not the average one. Select the gearbox so the motor spends most of its working life in an efficient operating region rather than near its thermal limit. Confirm that the controller and battery can continuously support the expected current instead of only surviving short bursts. Finally, validate every assumption through endurance testing because a drivetrain that performs consistently after hundreds of operating cycles is far more valuable than one that looks impressive during a single demonstration.
That sequence rarely produces the cheapest drivetrain, nor the most powerful one. It consistently produces something more valuable: a mobile robot that behaves the same on the thousandth cycle as it did on the first. That consistency is usually the difference between a successful product and a prototype that never quite earns the confidence of the engineers responsible for deploying it.
The content & opinions in this article are the author’s and do not necessarily represent the views of RoboticsTomorrow
Featured Product
