Collaborative Robots (Cobots) in Industrial Settings

Collaborative robots — universally abbreviated as cobots — represent a distinct class of industrial robot designed to operate alongside human workers within a shared workspace, rather than behind fixed barriers. This page covers how cobots are defined and classified under international safety standards, the mechanisms that enable safe human-robot interaction, the industrial scenarios where cobots deliver measurable value, and the decision boundaries that separate cobot deployments from conventional robotic automation. Understanding cobot capabilities is essential for any organization evaluating industrial robots in automation or planning a broader how industrial automation works conceptual overview.


Definition and scope

A collaborative robot is formally defined by ISO/TS 15066:2016 — published by the International Organization for Standardization — as a robot intended to physically interact with humans in a collaborative workspace (ISO/TS 15066:2016). This definition is operationally narrow: not every robot that shares a floor with a worker qualifies. The robot must be engineered and risk-assessed specifically for contact scenarios, distinguishing cobots from conventional industrial robots that happen to operate near humans.

The scope of cobots spans payload classes from under 3 kg to approximately 35 kg, with reach envelopes typically ranging from 500 mm to 1,300 mm. The dominant cobot form factor is a 6-axis articulated arm, though 4-axis and 7-axis variants exist for specialized tasks. ISO 10218-1 and ISO 10218-2 — the base standards for industrial robot safety — govern cobot installations alongside ISO/TS 15066, which adds collaborative-specific force and pressure limits.

Cobots are classified under the broader category of types of industrial automation as flexible automation assets: they are reprogrammable, physically relocatable, and capable of performing different task sequences without retooling the physical workspace.


How it works

Four distinct collaborative operating modes are defined in ISO/TS 15066, each enabling human-robot coexistence at a different level of proximity and physical contact:

  1. Safety-Rated Monitored Stop (SRMS): The robot halts when a human enters the collaborative workspace and resumes when the human exits. No contact is permitted during operation.
  2. Hand Guiding: A human operator physically moves the robot end-effector to teach positions or guide motion. Sensors at the flange detect applied force and translate it into robot motion.
  3. Speed and Separation Monitoring (SSM): The robot dynamically adjusts speed based on the real-time distance between itself and the nearest human. As a worker approaches, the robot slows; if the worker reaches a minimum separation threshold, the robot stops.
  4. Power and Force Limiting (PFL): The robot is designed to stop or retract upon detecting contact forces that exceed defined biomechanical thresholds. ISO/TS 15066 Annex A specifies body-region-specific force and pressure limits — for example, the fingertip limit is set at 30 N transient contact force and 140 N/cm² pressure.

PFL is the mode most commonly associated with cobot identity in the market. Robots implementing PFL use one of three hardware approaches: joint torque sensors, current-based torque estimation, or skin-type distributed tactile sensors on the outer surface.

Cobot motion control systems in industrial automation are architecturally similar to conventional industrial robots but with lower maximum joint speeds — typically 1,500 mm/s to 2,500 mm/s at maximum — and higher safety integrity levels (SIL 2 or SIL 3 per IEC 62061) in the safety-rated monitoring circuits.


Common scenarios

Cobot deployments are concentrated in tasks that combine variable positioning, light payloads, and direct human involvement at adjacent workstations.

Assembly: Cobots handle sub-component insertion, screw driving, and press-fit operations alongside human assemblers. Automotive interior trim lines and electronics PCB assembly are the two highest-volume deployment environments in US manufacturing, as documented by the Association for Advancing Automation (A3) in its annual robotics shipment data (A3 Robotics).

Machine tending: A cobot loads and unloads CNC machines, injection molding presses, or stamping equipment while a human operator performs setup, inspection, and changeover. This deployment pattern is particularly common in industrial automation for small and mid-sized manufacturers where floor space is constrained and workforce flexibility is required.

Quality inspection: Cobots position parts under fixed vision systems or carry machine vision and inspection systems end-effectors over complex surface geometries, reducing the ergonomic burden on human inspectors.

Palletizing and packaging: Light-payload cobots (under 10 kg) handle repetitive box placement and bag stacking at rates that relieve workers from repetitive strain injury risk without requiring the full guarding infrastructure of high-speed conventional palletizers.

Pharmaceutical and food processing: Regulated environments — covered in industrial automation in pharmaceuticals and industrial automation in food and beverage — use cobots where sanitizable end-effectors and frequent recipe changeovers make cage-guarded robots operationally impractical.


Decision boundaries

Cobots are not the correct choice for every automation problem. The decision to deploy a cobot versus a conventional industrial robot or a fixed vs flexible vs programmable automation system depends on four primary variables:

Factor Cobot appropriate Conventional robot appropriate
Payload Under ~35 kg Over 35 kg
Cycle time requirement Moderate (human-paced lines) High throughput, <15-second cycles
Human proximity requirement Mandatory collaboration Separation feasible
Product mix High-mix, frequent changeover Low-mix, high-volume, stable process

Payload ceiling is the hardest constraint. ISO/TS 15066 force limits restrict the kinetic energy permissible at contact, which mathematically limits safe operating speeds for any given payload. Exceeding approximately 35 kg at cobot speeds violates the biomechanical thresholds in Annex A unless the workspace is fully fenced during robot motion, at which point collaborative operation is no longer occurring.

Cycle time represents the second hard boundary. Cobot operating speeds average 40%–60% of the maximum speeds of comparable conventional industrial robots. Applications requiring sub-10-second cycle times across high volumes — such as industrial automation in automotive body welding — cannot be served by cobot-class systems without accepting throughput penalties.

Integration complexity should account for the full industrial automation system integration burden: cobot deployments typically require less guarding infrastructure but more sophisticated end-of-arm tooling (EOAT) design and more iterative force-threshold calibration than cage-guarded robots. The workforce implications are addressed in industrial automation workforce impact, where cobot deployments show a different redeployment pattern compared to full automation cells.

For organizations at the evaluation stage, the industrial automation implementation lifecycle provides the structured framework for moving from task analysis through risk assessment — including the mandatory collaborative workspace risk assessment required by ISO/TS 15066 Section 5 — to validated deployment.


References