The modern manufacturing floor is undergoing a quiet revolution, not through full automation, but through collaboration. Industrial robotics, once confined to safety cages performing repetitive, high-speed tasks, is being transformed by the advent of collaborative robots, or cobots.
Author: Samantha Carroll
25/11/23
This paper examines the integration of cobots into existing manufacturing lines, focusing on the triad of safety standards, advanced perception, and the design of effective human-robot interaction (HRI) to achieve true synergy.
Unlike traditional industrial robots that operate in isolated workspaces, cobots are designed to share a workspace with human workers, necessitating a fundamental redesign of safety paradigms.
We begin with an analysis of the international safety standard ISO 10218 and its companion specification ISO/TS 15066, which defines the requirements for collaborative robot operation. It outlines four types of collaboration: safety-rated monitored stop, hand guiding, speed and separation monitoring, and power and force limiting (PFL).
PFL is the most interactive form, where the cobot's inherent design (rounded edges, padded surfaces) and embedded force-torque sensors ensure that any unintended contact with a human results in forces below pain thresholds. Implementing these standards requires a risk assessment for each collaborative task, considering factors like robot speed, payload, and the human body part at risk. Beyond passive safety, active perception is key. We detail the sensor suites enabling cobots to work intelligently alongside people.
This includes 2D and 3D vision systems for part identification and pose estimation, area scanners (e.g., laser scanners) that create dynamic protective fields around the robot, slowing it down as a human approaches and stopping it upon intrusion, and the critical role of force-torque sensing at the wrist.
These sensors enable advanced capabilities like precise force-controlled assembly—inserting a bearing into a housing with sub-millimeter accuracy by feeling the forces—and programming by demonstration, where a human operator physically guides the cobot through a task, which it then repeats autonomously.
The final, often underestimated component is HRI design. Effective collaboration requires intuitive communication interfaces. This can be through simple LED light bars indicating robot status (idle, moving, in error), touchscreen panels with clear graphical workflows, or even natural language commands for simple instructions. The workspace must be ergonomically designed for both human and robot, considering reach, part presentation, and clear sightlines. A case study from automotive electronics assembly illustrates this integration.
Here, a cobot performs the precise, ergonomically taxing task of applying thermal paste to circuit boards, while a human operator performs the subsequent, dexterity-demanding component placement and visual inspection. The cobot's speed and separation monitoring system ensures it halts its arm movement if the operator leans too far into the shared cell. This synergy increases overall line throughput by 30% while reducing repetitive strain injuries for the human worker.
We conclude that successful cobot integration is a systems engineering challenge that equally prioritizes rigorous safety compliance, sophisticated environmental awareness, and thoughtful human-centric design to unlock new levels of productivity and flexibility in manufacturing.