Robot Design Around the World: How Form Follows the Job

A robot arm built for welding does not need to look like a humanoid. A warehouse mobile robot does not need a face. A surgical robot, drone, cobot and factory arm solve different physical problems. Their shapes are not cosmetic choices. They are engineering answers to task, environment and cost.

That is why a global design map is useful. It shows how wide the field already is. Robotics is not waiting for humanoids to become useful. Industrial arms, AMRs, inspection robots and specialized service systems have been doing physical work with forms that fit their jobs.

The post names ABB's IRB Series and KUKA's KR CYBERTECH Series as industrial arm examples. These systems are designed around repeatability, reach, payload, duty cycle, safety integration and factory uptime. Their value comes from doing the same motion accurately for long periods.

That is a different design culture from humanoids. A humanoid has to move through spaces made for people. An industrial arm can be installed inside a controlled cell. The constraints are different, so the bodies are different.

Humanoids attract attention because their body plan suggests generality. Legs, arms and hands make sense in human spaces. The cost is complexity. Balance, manipulation, batteries, perception, safety and heat all become harder when the robot has to carry itself through the world.

A strong robot design map should not crown one form factor. The useful comparison is task fit: required reach, payload, mobility, tool access, installation space, safety boundary and maintenance burden.

ABB's current catalog ranges from small IRB arms to larger industrial families and AMRs. The IRB 1200 Next Generation, for example, is optimized for high-speed, high-precision tasks with variants up to 9 kg payload, while larger ABB families cover much heavier loads.

KUKA's KR CYBERTECH family uses the same principle of task-specific variants. Current models span multiple payload and reach combinations and support handling, assembly, machining and welding workflows. These are mature examples of designing the body around a constrained process instead of asking one robot morphology to do everything.

Start with the work envelope. Record the horizontal and vertical reach needed at every tool pose, then add clearance for fixtures, guards and cable routing. A six-axis arm can cover a large fixed volume, while a mobile base trades mounting rigidity for access to several stations. A humanoid adds stairs, doorways and human-height work surfaces, but every moving joint adds mass, wiring, control effort and service points.

Next compare payload at reach, rather than payload as one headline number. A gripper, camera, force sensor and tool all consume payload before the workpiece is lifted. For mobile robots, include floor transitions, turning radius, slope limits and localization conditions. For legged systems, include balance margin and the effect of carrying a load away from the torso.

Record how the robot fails and how a technician recovers it. Check emergency-stop architecture, guarding or speed-and-separation controls, tool-change access, calibration steps, replacement time for wear parts and whether a failed joint blocks the full cell. These details help determine whether a body shape fits a production task.