5 stages of human/robot interaction

Features - Robotics

Stäubli North America’s Sebastien Schmitt discusses automation integration from fully enclosed systems to collaborative robots.


With any automation project, industrial designers need to balance cost with performance, flexibility with speed, and safety with productivity. Modern sensing and control systems allow robots to work alongside employees for maximum flexibility or completely walled away from people for increased safety. Robot producer Stäubli has defined five levels of man-robot collaboration (MRC) and introduced the TX2 robot with its CS9 controller to support each level. The robot-controller combination allows manufacturers to use the same hardware for fully enclosed to fully collaborative robots. Robotics Division Manager Sebastien Schmitt recently sat down with Today’s Motor Vehicles to discuss the pros and cons of interaction stages, and how to configure systems.

Stage 1

Hard guarding separates robot from operators; manufacturing process performed entirely by robots

Pros: Robots can safely operate at highest possible speeds; simple safety systems
Cons: Lowest flexibility; interacting with robots requires full machine stoppage

Schmitt: “Full speed is possible when the robot is fenced in. High productivity comes from being able to take advantage of the full performance of the robot. You can run at the maximum speed of the robot.”

There are fewer safety concerns when robots are completely enclosed. When the door to the cage opens, robots stop moving.

“As you move toward collaboration, the more in-depth risk analysis you have to do. If people aren’t interacting with robots, safety is much easier to evaluate.”

Stage 2

Laser, virtual guarding separates robot from operators; manufacturing process performed entirely by robots, human operators enter area periodically

Pros: Supports palletized production, more flexible than Stage 1
Cons: More interaction with operators creates safety challenges; can be slower than Stage 1

Schmitt: “Instead of being completely enclosed, you open a side of the system, so you open up more risk by creating the ability for someone to walk in and walk out.”

Typically, users set up laser scanners and safety curtains to determine when a person is near the robot and program controls to react accordingly. Multiple systems create redundant backups to ensure that robots are not moving at full speed when people are in proximity.

Most users who opt for Stage 2 instead of Stage 1 MRC do so for material loading and unloading.

“If you have a box coming in from a conveyor and you want operators to come in and unload pallets when they’re full, you want robots to be able to continue working,” Schmitt says. “You don’t want a door to open that will stop the robot. You don’t want the operator to have to push a button and stop the system. They need to interact with the environment.”

Stage 3

Laser, virtual guarding separates robot from operators; robots and operators share manufacturing processes; people regularly enter work zone

Pros: More flexible than Stage 2; faster than collaborative environments
Cons: Complex safety requirements; slower than Stages 1, 2

Schmitt: “Stage 3 requires the operator in its process. In Stage 2, the machine is autonomous, but in Stage 3, the user is closer to the machine more often.”

The mid-point in MRC allows for great flexibility in system configuration because the operator will direct workflows and perform some process or assembly operations. The robot can still operate quickly, but users have to go through complex safety planning to ensure that the robot is aware of where the operator is at all times and adjust its operations accordingly.

“In using a robot in a collaborative way, you have to be cautious not to have sharp angles as part of the design because those create risk to the operator,” Schmitt says. “The zones have to be properly defined. You have to very carefully balance throughput, speed, and safety. It’s a tradeoff. The speed required for high-volume operations are not compatible with collaborative environments.”

Stage 4

No physical separation between operators and robots; robots and operators share manufacturing processes; robots stop when they contact people

Pros: Highly flexible; operator can adapt manufacturing processes
Safety requirements limit robot force, speed; robotic safety skin needed

Schmitt: “The robot is moving while the person is standing next to it, so we need to be able to interact with it and stop the system. We put a safety skin on the robots that we call TX2 touch that senses any contact with the operator. If it senses any contact, the robot stops.”

Because people and robots are operating in the same space, the system must limit robotic capabilities while operators are present. So, Stage 4 systems are slower than earlier stages, but they can still be attractive for manufacturers hoping to boost flexibility.

“You can still work the robot at full speed, but it only does that when no one is standing next to it,” Schmitt says, adding that proper configuration of laser scanners and light curtains is critical. “When a person enters the zone near the robot, the machine slows to a safer mode.”

Stage 5

No physical separation between operators and robots; robots and operators share manufacturing processes; humans touch and move robots, directing their motion

Pros: Full collaboration between operator and robot; most-flexible option; no safety skin required
Cons: Slowest option; special controls needed for man-machine interaction

Schmitt: “The person uses the robot as a tool to help perform an operation in this configuration. With fine assembly, for example, a person can hand-guide the robot to pick up a part and move it to another part for assembly. For heavy pieces, a person couldn’t hold both pieces for assembly, so the robot handles the weight, but the person hand-guides the position.”

Rather than use a skin to sense contact, the robot expects the person to touch systems. So instead of stopping when it senses contact, the robot stops when the person is not touching the controls. Operators interact with machines using controls with force sensors – when the person pushes the control in a direction, the force sensor interprets that motion and moves the robotic arm accordingly. When the operator isn’t touching the controls, the robot does nothing.

Stäubli North America

About the author: Robert Schoenberger is the editor of TMV and can be reached at 216.393.0271 or rschoenberger@gie.net.