When most people think about AI, they picture something that happens on a screen — chatbots answering questions, software generating images, or algorithms analysing data. But one of the most exciting developments in artificial intelligence is happening off-screen, in the physical world.
A new research paper titled “SPINE: Bridging the Cyber-Physical Gap with Agentic AI” shows how AI is learning to control robots and perform physical tasks, opening up possibilities for Irish manufacturing, construction, and trades businesses that go far beyond anything a chatbot can do.
The Problem with Getting AI into the Real World
Giving AI a body has been one of the hardest challenges in robotics. While AI systems have become incredibly good at thinking — planning, analysing, and making decisions — translating that intelligence into physical action has required teams of expert engineers spending weeks or months on calibration.
Every robot is different. Every factory floor is different. Every task — welding a joint, packing a box, sorting components — requires precise adjustments that only skilled technicians could handle. This “deployment gap” has been the main bottleneck preventing AI-powered robotics from becoming practical for smaller businesses.
The SPINE research tackles this problem head-on. The researchers built an AI system that can take a robot out of the box and get it working with minimal human expertise. In their experiments, a complete novice using SPINE outperformed experienced human operators who were using standard AI tools but without SPINE’s structured approach.
What the Research Found
The results were striking. With SPINE, operationalisation success improved from 75% to 100%. The time needed to get a robot working dropped from an average of 16 minutes and 45 seconds to 13 minutes and 47 seconds. More importantly, the system worked across different types of robots — it was tested on DOBOT and AgileX platforms — showing it isn’t limited to one specific machine.
The researchers also implanted bugs in the robots to see how well the system could diagnose and fix problems. SPINE resolved all 10 implanted bugs, matching the performance of expert human technicians. This means that AI can now handle troubleshooting that previously required years of specialised training.
What This Means for Irish Businesses
For Irish small and medium businesses, this is potentially transformative. Ireland has a strong manufacturing sector, from food processing and pharmaceuticals to construction materials and precision engineering. But robotics has traditionally been out of reach for smaller operators because of the expertise required to set up and maintain automated systems.
Imagine a small Irish manufacturing company that wants to automate a packaging line. Currently, they would need to hire a robotics engineer to configure the system — a scarce and expensive resource. With AI-powered systems like SPINE, a production manager with basic technical skills could get the robot up and running in an afternoon.
For trades businesses — welding, fabrication, assembly — this means AI-powered robotic assistants that can be deployed quickly and reconfigured for different tasks without calling in specialist engineers every time.
Practical Next Steps
How can an Irish business prepare for this shift? Here are four practical suggestions:
1. Start learning about what’s available. Irish companies like Forge Robotics and Horan Automation are already active in this space. Familiarise yourself with what AI-powered automation can do.
2. Identify repetitive physical tasks. Look for tasks in your operation that are repetitive, physically demanding, or require high precision. These are the best candidates for AI-powered automation.
3. Plan for a skills shift. The expertise needed to deploy robotics is changing from specialist programming to general technical competence. Your existing team members may be able to operate these systems with relatively short training.
4. Start small. You don’t need to automate your entire factory. Pick one task, find a system that can handle it, and learn from the experience before scaling up.
The gap between AI’s brain and its body is closing. For Irish businesses in manufacturing, construction, and trades, that means a future where powerful automation is no longer reserved for the largest companies with the deepest pockets.