BUSINESS MATTERS
and operational disruption. Yes – information inputted into free
AI models can become public training data, leading to corporate secrets or personal data being exposed. However, these are not the AI systems that are going to add value in industrial settings. Keeping it local and using internal GPU
(Graphics Processing Units) and servers to run AI models offers maximum data privacy, ensuring that no sensitive data is used for external training or leaked to third-party cloud providers. On-device AI is also a high security option, as the data is processed on user devices rather than in the cloud. Where companies are using established cloud-based platforms such as Microsoft Azure or Google Cloud Platform for their AI applications, they can be sure their data is safe too, as both offer robust security. However, using a lesser known supplier, without thorough research, could be a risky move. Security does become a potential issue
when data is being shared with a third party such as a machine supplier, for diagnostics, troubleshooting and servicing. There are best practices that can be adopted for secure sharing, such as processing data through automated checkers to remove sensitive information. Ultimately though, it is a case of working with trusted suppliers, and of controlling what data goes into the AI system in the first place. Going forwards, AI systems are
only going to become more secure as legislation catches up with technological advancement. In 2027, the EU Cyber Resilience Act (CRA) comes into force, ensuring mandatory cybersecurity requirements on all software and hardware. AI products will have to be secure by design, with no known
exploitable vulnerabilities, undergo risk assessments and offer lifecycle support, including security updates. The new Fanuc R50iA controller – based
on the latest CNC hardware, FS500 – is already fully CRA compliant. Over the next two years, all Fanuc robots will feature this new controller as standard, while the oIder controllers already provide a very high level of security. Fanuc customers can rest assured that they will be ahead of the cyber resilience curve, with their machines futureproofed thanks to smart, safe and compliant equipment.
Myth #3: AI is moving so fast systems will become obsolete very quickly The pace of development in AI is rapid, but there are certain measures that companies can take to avoid obsolescence. Open ecosystems and standardised, open- source platforms are a crucial aspect of futureproofing, particularly with the shift from generative to agentic AI. MCP (an open-source protocol that
allows AI assistants to interact with external data sources and software tools) is one standard interface of such systems. OPC Unified Architecture (OPC UA) is another secure, open-source and platform-independent IEC62541 standard for industrial communication.
Fanuc is actively collaborating with
AI-trailblazer NVIDIA to bring physical AI into mainstream manufacturing. A key step in this journey is our support for the open-source robotics platform ROS 2, which enables programming via Python. By lowering the barrier to entry, this allows developers, researchers and companies to build AI-driven robotics applications on Fanuc’s proven industrial hardware.
Myth #4: AI is just a gimmick – it won’t really improve our factory automation AI can add value to production lines in more ways than we can currently imagine. Already, predictive maintenance is using sensor data to forecast equipment failures before they happen, vision systems are detecting defects in non- uniform products and diagnostics are dramatically reducing downtime. AI is transforming industrial
automation too, making robots smarter, safer and faster to deploy, through voice-controlled operation, adaptive motion control, safety-aware human-robot collaboration, and virtual commissioning via digital twins. One of the most significant benefits of
AI is its ability to accelerate deployment. By assisting with code generation, AI makes it easier and quicker for companies to implement robotic systems. AI-enabled robots also allow existing production lines to be retrofitted without extensive modifications, further speeding up the rollout of automation.
Myth #5: We’re not AI-ready – it’s going to cost us a fortune To use AI in manufacturing, a company does not need to be fully digitised, but it does need a solid data foundation, as ultimately, AI without data is pointless. AI requires data to be structured, contextualised and available in real- time to drive applications like predictive maintenance, quality control and line optimisation. While not every machine needs to be ‘smart’, critical equipment requires sensors to monitor performance and feed data to digital twins. Regarding cost, one of the benefits of AI
is that it is quantifiable – companies should be able to understand what their return on investment (ROI) is going to be from the start and gauge at a very early stage whether a project is going to work or not. That said, the more deeply companies
commit to AI, the more they will likely get out of it. McKinsey reported that organisations with the most ambitious AI agendas are seeing the greatest benefit. These firms are often aiming to achieve more than just cost reductions; their objectives extend to driving growth and/or innovation through AI, which is where its real potential lies.
www.fanuc.eu/uk/en www.welding-world.com WeldingWorld1 / 33
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40
