Shaping the future of safe automation

The International Robotic Safety Conference (IRSC), hosted by the Association for Advancing Automation (A3), took place in Houston, Texas, from November 3 to 5, 2025. As Clara Sipes, market analyst at Interact Analysis explains, it highlighted how safety requirements are evolving within one of the fastest-growing segments of industrial automation.

With 28 speakers, 20+ exhibitors, and 100+ attendees, the event was able to cover many relevant topics, including:

• Updates to the ISO 10218 (industrial robots) international safety standard
• Humanoids, four-legged robots, and wheeled AMRs
• Functional Safety
• The Cyber Resilience Act, AI Act, and Machinery Regulations

These topics effectively cover how the safety and standards communities are addressing new applications, such as humanoids or collaborative robots, to ensure the technologies can be deployed safely and at scale in the future.

Humanoids and legged robotics gain attention

Humanoid and legged robots were a major discussion point at the conference, in anticipation of their mass deployment in the future. The unique mechanics associated with legged robotics introduce new challenges around stability, fall dynamics, and unpredictable motion.

New standards such as ISO 26058-1 and ISO 25785-1 have been developed to address both statically and dynamically stable mobile robotics, acknowledging the unique residual risks associated with these platforms. Unlike statically stable machines, dynamically stable machines, such as humanoids, collapse when power is cut, creating residual risk in the event of a fall.

Early progress in this area is also captured in TR R15.108, which examines the challenges associated with bipedal, quadrupedal, and wheeled balancing mobile robots. Beginning work on this standard now ensures guidance will be in place as the technology scales up.

Key updates to ISO 10218

A full session of the IRSC was dedicated to showcasing revisions to ISO 10218-1 (industrial robots) and 10218-2 (industrial robot applications and robot cells).

The main changes centre on new developments that are missing from existing standards: risk-based collaboration, explicit functional safety requirements, structured lifecycle coverage, and expanded documentation obligations.

One of the most notable shifts is the removal of references to “collaborative modes.” According to the revision, collaborative robots must be evaluated based on the application, not the robot alone. Each application carries its own risks, and the standard now encourages assessing the entire environment within which the robot operates.

For manufacturers, these changes will influence product development, documentation, and functional-safety architecture. For integrators, the emphasis on robot cell safety will raise the importance of comprehensive safety being implemented through risk assessments.

Functional safety is becoming a system-level expectation

Functional safety was emphasised across multiple sessions, especially as safety functionality is increasingly embedded within motors, drives, and controllers. Standards such as ISO 13849 and ISO 62061 remain the foundation for integrating functional safety into industrial robotics. Important themes included:

• Evaluating Safety-Related Parts of Control Systems (SRP/CS)
• The growing prevalence of functional safety in robotic cells
• Designing safety at the input, logic, and output levels
• Understanding Safety Integrity Levels (SIL) within specific safety functions

For motor and drive manufacturers, this reinforces the trend toward integrating functional safety capabilities to support a comprehensively safe system.

Cyber Resilience Act, AI Act, and Machinery Regulation

European regulatory changes, particularly the Cyber Resilience Act (CRA), AI Act, and Machinery Regulation, were on display at IRSC. They are shaping the future of industrial automation by establishing a unified framework for safety, cybersecurity, and risk management.

The primary theme of these regulations is new vulnerabilities within industrial automation products, as they are increasingly connected to a network. In order to comply, products will have to be designed with lifecycle-long cyber resilience in mind, implement structured risk assessments for AI functions, and meet updated machinery compliance requirements that now formally incorporate both AI and cybersecurity considerations. As these rules come into effect, suppliers and machine builders will face greater expectations for documentation, validation, and ongoing support. It will be particularly apparent when the CRA and Machinery Regulation become mandatory in 2027. This will ultimately create a more integrated and rigorous regulatory ecosystem governing how digital and AI-enabled technologies are developed and deployed across industrial automation.

How will regulatory changes impact the market?

We spoke with industrial robot vendors about the potential impact of the upcoming ISO 10218 standard on the robotics market. Most companies interviewed believe it will significantly affect R&D efforts, with some expecting increased production costs due to new requirements. However, they noted that this challenge will be shared across the industry, as all manufacturers must comply. As the new standard is set to take effect in Europe in 2027 and later in the US, many robot vendors in these regions are already developing solutions to meet the new standard. Emerging robot suppliers from Asia have also started studying the new standards, and some have developed their solutions for the European market, while others remain in a wait-and-see phase.

With the adoption of the term “collaborative applications” instead of “collaborative robots (cobots)” in the new standard, we anticipate that cobots will continue to penetrate more traditional industrial robot applications. These include palletising, welding and inspection – all of which have seen rapid growth in the past two years. In 2025, cobots are expected to account for just under 10% of global industrial robot shipments, but the share is projected to rise to 16% by 2030 in our latest Industrial Robots report.

In the humanoid robot space, the lack of safety regulations and technical standards is one of the most frequently mentioned barriers to large-scale adoption. At present, when users deploy humanoid robots – especially bipedal robots – they must bear the risks of personal injury and equipment damage caused by robot falls. An established standard system for dynamically stable robots, including standards such as ISO 26058-1 and ISO 25785-1, will effectively alleviate such concerns. Furthermore, improved and referenceable standards will help manufacturers accelerate technical iteration and expand production scales.

Beyond the mobile stability of hardware, the strong correlation between humanoid robots and AI means that supervision at the AI level is equally crucial. Data collection and model training are standard processes that AI-driven humanoid robots typically undergo before real-world deployment, so the protection mechanism for private domain data needs to be emphasized. After deployment, robot operations based on AI algorithms may also trigger user concerns about cybersecurity. Therefore, the improvement of AI regulations also contributes to the healthy development of humanoid robots.

What should manufacturers and integrators do now?

Prepare early for upcoming standards revisions: With multiple regulations taking effect over the next few years, organizations that begin aligning now will avoid costly redesigns and rushed compliance efforts later. Early preparation allows manufacturers to phase in new documentation practices, update product development roadmaps, and adjust gradually rather than reactively. This proactive approach not only reduces risk but also positions companies to bring compliant products to market faster than competitors who wait.

Shift toward built-in safety functionality to meet future requirements: Safety functionality will begin to be seen as a necessity, rather than as a convenient add-on. Manufacturers that begin considering safety in the design process now will have a competitive advantage in the approval process. Customers are also aware of the upcoming changes in regulations and are more likely to purchase products designed with them in mind.

Strengthen cybersecurity and AI compliance processes: As more devices are connected to a network and AI becomes more prevalent within industrial automation products, cybersecurity and AI will play a pivotal role. Companies will have to restructure, with teams established to focus on delivering the necessary updates to comply with standards. Additionally, documentation processes should be established now so that staff can be trained with compliance in mind. This will ultimately reduce exposure to cyber incidents while also ensuring compliance with CRA, the AI Act, and the Machinery Regulation.

In conclusion, the International Robotic Safety Conference highlighted how rapidly the safety environment is evolving, driven by new standards that reflect innovative robotics technologies and tightening global regulations. Staying ahead of standards, investing in safety architectures, and preparing for new compliance demands will be key to ensuring safe, scalable robotics deployment in the years ahead.