The latest in sensors, programming, and risk assessment for safe collaborative deployment
Collaborative robots – or cobots – are rapidly transforming industrial automation by working safely alongside human workers.
Unlike traditional industrial robots, which are typically caged off, cobots are designed for direct interaction, enabling flexible, efficient workflows.
But as adoption grows, so too does the need for advanced safety systems, smarter programming environments, and comprehensive risk management frameworks.
This article explores the latest innovations in human-robot collaboration, focusing on how evolving safety standards, sensor technologies, and software advances are enabling the next generation of safe and scalable cobot deployments.
The changing regulatory landscape
Cobot safety is defined by a range of international standards, most notably ISO 10218 and ISO/TS 15066, which together outline the key requirements for robot system safety and human-robot collaboration.
These standards continue to evolve to reflect new use cases and technological advancements.
While ISO 10218 governs industrial robot safety more broadly, ISO/TS 15066 is the cornerstone for collaborative applications, defining thresholds for permissible contact forces, types of collaboration (for example, hand-guiding, speed and separation monitoring), and the required protective measures.
More recently, ISO/TS 15066 has served as the foundation for newer guidelines such as RIA TR R15.806 and EN ISO 13849, which provide additional clarification on functional safety and performance levels for safety-related parts of control systems.
These standards are pushing manufacturers and integrators to design for safety at both the hardware and software levels.
Advanced safety sensors: From vision to proximity
The backbone of cobot safety lies in real-time environmental awareness. Key sensor innovations are enabling machines to detect and respond to human presence with ever greater speed and accuracy.
1. Vision-based systems
AI-powered 3D vision cameras are now capable of differentiating between people, objects, and background environments. These systems support dynamic zone mapping, allowing cobots to slow down, stop, or reroute when a person enters a shared workspace.
2. Force-torque sensors
Built-in at the robot’s joints or end-effectors, these sensors enable tactile awareness. If a cobot encounters unexpected resistance or impact, it can react immediately – either by stopping or adjusting its motion path.
3. Proximity and ultrasonic sensors
These sensors create virtual safety zones, similar to automotive collision-avoidance systems. They’re often used in combination with vision systems for layered safety coverage.
4. Wearable safety devices
Some systems now include human-worn transponders or haptic belts that alert operators of unsafe proximity or provide feedback when in shared zones with a robot.
Smarter programming for safer collaboration
Low-code and no-code programming environments are empowering non-experts to configure cobots while embedding safety constraints from the start.
Graphical safety programming tools allow integrators to define safety zones, velocity limits, and toolpaths with visual confirmation. These platforms integrate with digital twins, enabling virtual testing of safety responses before deployment.
Teach pendant systems – now often equipped with safety override features and soft stops – allow operators to guide robots through tasks physically, ensuring intuitive and inherently cautious path planning.
Additionally, AI and machine learning are starting to play a role in dynamic safety adaptation. By learning from historical data and human feedback, some cobots can autonomously adjust speed, path, or tool pressure depending on context and task variation.
Risk assessment: A critical pillar of safe deployment
No cobot deployment should proceed without a comprehensive risk assessment, as mandated by ISO/TS 15066 and EN ISO 12100. This involves identifying potential hazards, estimating their severity and likelihood, and implementing mitigation strategies.
Key elements of a robust cobot risk assessment include:
- Task analysis: Understanding the physical interaction required, frequency of human involvement, and complexity of the motion.
- Workspace mapping: Defining collaborative zones, restricted areas, and safe egress paths.
- Force and pressure limits: Ensuring that any potential contact remains within human-safe thresholds.
- Emergency response planning: Including redundant stop buttons, safe robot recovery protocols, and training for all staff.
Modern software tools such as SISTEMA (developed by IFA Germany) can aid in calculating safety performance levels based on system architecture and intended use.
10 leading companies providing safety systems for cobots
Here are some of the top technology providers developing components and systems for safe human-robot collaboration:
- Sick – Market leader in vision-based safety systems and LiDAR for robots.
- Pilz – Specializes in programmable safety controllers and modular safety relays.
- Keyence – Offers compact, high-resolution safety sensors and zone scanners.
- Cognex – Advanced machine vision and AI-based safety systems.
- ABB – Its SafeMove and OmniCore platforms combine motion control with safety.
- Universal Robots – Built-in force limiting and user-friendly safety configurations.
- Kuka – Offers LBR iiwa cobots with torque sensors and integrated safety features.
- OnRobot – End-effectors with built-in sensing and compliant gripping.
- Fanuc – Provides Dual Check Safety (DCS) with customizable safety parameters.
- Omron – Integrated safety platforms with adaptive motion and vision systems.
A safer, smarter cobot future
As cobots become increasingly common in warehouses, factories, and labs, safety is no longer a bolt-on feature – it’s a foundational element.
With advances in sensors, intuitive programming, and international standards pushing for harmonized best practices, the future of human-robot collaboration is both promising and secure.
Yet the onus remains on integrators, manufacturers, and end users to ensure that deployment is done thoughtfully. Safety isn’t just a technical specification – it’s a shared responsibility.
