ROBOTIC.INDUSTRIES

Industrial Robotics

Payload vs Reach: Why a Bigger Robot Arm Gets Slower

Payload and reach fight each other through joint torque. Here is the arithmetic, the derating you should expect at full extension, and how to size an arm without buying twice.

Six-axis industrial robot arm in motion inside a guarded manufacturing cell
Six-axis industrial robot arm in motion inside a guarded manufacturing cell

Reach costs payload because a robot joint is limited by torque, and torque is force times distance. Doubling the reach of an otherwise identical arm roughly halves what it can carry at the same joint torque, and it also lowers the achievable acceleration, because the moment of inertia of the arm rises with the square of length. That is why a 1,800 mm arm is never a longer version of a 900 mm arm; it is a heavier, slower and more expensive machine.

how inertia scales with reach
30 to 50 %payload lost to a long centre of gravity offset
3xarm mass increase from 5 kg to 20 kg class
80 %of rated payload, the practical working ceiling

The three quantities that fight

Every axis has a torque budget set by its motor and gearbox. Three demands compete for it.

  • Static load. Payload mass times gravity times the horizontal distance from the joint. A 10 kg payload at 1.5 m produces about 147 Nm at the shoulder before anything moves.
  • Dynamic load. Inertia times angular acceleration. Inertia grows with the square of the distance, so moving the same mass twice as far out demands four times the torque for the same acceleration.
  • Self weight. The arm's own links, motors and gearboxes. On long arms this is the dominant term, and it is the reason a 20 kg-class arm at 1,750 mm weighs 62 kg to 68 kg while a 16 kg-class arm at 950 mm weighs about half that.

The consequence is a rated payload that is honest only under stated conditions: a defined centre of gravity offset, usually 50 mm to 100 mm from the flange, at rated speed and duty. Every real installation violates at least one of them.

What the datasheet does not print

Effects that reduce usable payload below the rating
EffectTypical lossWhy it happens
Centre of gravity moved 60 to 200 mm off flange30 to 50 %Wrist joints see torque, not mass
Operation at full horizontal extension10 to 25 %Shoulder and elbow torque peak
High acceleration profile15 to 30 %Inertial term dominates static load
Wrist rotated to a torque-unfavourable pose10 to 40 %Axis 5 and 6 are the weakest joints
High duty cycle at elevated ambient5 to 15 %Motor and gearbox thermal limits
Collaborative operationvariesEffective mass caps permissible contact speed

These stack. A 10 kg-rated arm carrying a gripper whose centre of gravity sits 220 mm from the flange, at full extension, on an aggressive profile, can be genuinely limited to 4 kg to 5 kg of part. The controller will usually let you try, then report a torque or following error at the worst point in the path, three weeks after commissioning.

Load data is not paperwork. Entering the true mass, centre of gravity and inertia tensor of the tool lets the controller plan a feasible trajectory. Leaving the default values in place is the most common cause of unexplained path deviation and of gearbox wear that appears years early.

How manufacturers actually resolve the trade

Instead of one scalable design, builders offer families that sit at different points on the curve. Reading a range this way makes selection much faster.

Typical industrial arm families and their trade point
PayloadReach bandArm massDesign emphasis
3 to 7 kg600 to 900 mm25 to 55 kgSpeed and repeatability
10 to 20 kg1,400 to 1,800 mm150 to 260 kgGeneral handling, welding
50 to 80 kg2,000 to 2,300 mm550 to 750 kgSpot welding, machine tending
120 to 240 kg2,600 to 3,100 mm1,100 to 1,700 kgPalletising, press tending
500 to 1,000 kg3,100 to 4,700 mm4,500 to 8,000 kgBody handling, foundry
above 1,500 kg3,000 to 3,500 mmabove 10,000 kgHeavy assembly, rail and aerospace

Note how arm mass grows far faster than payload. Going from the 20 kg class to the 240 kg class multiplies payload by 12 and arm mass by roughly 7, and the foundation, the floor loading and the cost of moving the machine grow with it.

A sizing method that does not fail late

  1. Sum part, gripper, coupler and anything the gripper carries incidentally.
  2. Measure or estimate the centre of gravity offset from the flange face. If it is above 150 mm, expect to need the next class.
  3. Divide the total by 0.8 to leave headroom. That is the rated payload you buy.
  4. Check reach from the mounting point to the furthest work position, then add 10 % for approach and retract vectors that are not straight lines.
  5. Verify the worst pose, not the average one. Full extension with the wrist rotated is where torque peaks.
  6. Run the manufacturer's sizing tool with the real inertia tensor before ordering.

Step 4 catches a classic mistake. A cell drawn with the robot 1,250 mm from the far pick point looks fine against a 1,300 mm arm, until the approach vector requires the tool to stand off 120 mm and the arm cannot reach the position in a feasible pose.

Frequently asked questions

Why does a longer robot carry less?

Because joints are torque-limited. Static torque rises linearly with distance and inertia rises with the square of it, so the same motor and gearbox deliver less usable payload as reach grows, and the arm's own mass consumes a larger share of the budget.

How much payload does the centre of gravity offset cost?

Moving the load centre of gravity from about 60 mm to 200 mm off the flange typically costs 30 % to 50 % of rated payload, because the wrist axes are torque-limited rather than mass-limited. Long vacuum bars and offset nests are the usual cause.

Can I run at 100 % of rated payload?

Only under the datasheet conditions. In practice, size to about 80 % of rating so that speed, duty cycle and an unfavourable pose still fit inside the torque budget.

Does reach affect repeatability?

Yes. Repeatability is quoted at a defined test pose, and longer arms generally publish a larger figure because joint angle error is amplified by link length. Compare like with like, and check the test condition, not just the number.

What if I need long reach and high payload?

Move the robot instead of extending it. A shorter, stiffer arm on a linear seventh axis usually delivers better accuracy, higher speed and lower cost than a very long arm sized for the same envelope.

Sources

  1. ISO 9283, manipulating industrial robots, performance criteria and test methodsDefines the conditions under which payload, pose accuracy and repeatability are stated
  2. ISO 10218-1:2025, Robotics, Safety requirements, Part 1International Organization for Standardization, definitions of rated payload and workspace
  3. World Robotics 2025, industrial robotsInternational Federation of Robotics, application mix by payload class