Exeter-based
Rapid Fusion has unveiled ‘Bob’ (Base of Build), a new artificial intelligence (AI) powered ‘3-D print assistant’ that will help enhance the performance of its groundbreaking robotic additive manufacturing (AM) systems. The South-West based technology specialist’s team of developers have leveraged the power of AI to optimise solutions for customers 3-D printing complex components for the automotive, aerospace, construction, medical and military sectors.
After eight months of coding, troubleshooting and various live tests, Bob is now being rolled out to existing customers and is promising to optimise its robots by making them easier to use, providing greater operator control and ensuring less downtime through preventative maintenance.
The pre-loaded knowledge bank, anticipated to be one of the largest collections of 3-D printing expertise ever created, works in both secure online/cloud-connected and offline/air-gapped configurations for military or IP-sensitive customers. This works with Rapid Fusion’s existing models like Apollo and Zeus and for its most recent system Medusa, the first UK-built large-format hybrid 3-D gantry printer.
Martin Jewell, Rapid Fusion’s chief technology officer, said: “There is a lot of talk about how AI can transform the business world and numerous companies are jumping on the bandwagon. It is something we have been aware of from day one and all of our robotic AM systems have been built so that we can harness AI to unleash the full potential of our technology.
Boosting efficiency“Having our own ‘AI print assistant’ is a game-changer and will cut machine downtime and boost efficiency. We are teaching our systems to understand challenges and different scenarios, which means we can make the user interface more responsive and simpler to embrace — opening it up to all the workforce. In essence, if we can make our systems as ‘plug and play’ as possible it means we will have more adopters. It is a simple as that.”
Rapid Fusion, which employs 10 people at its demonsration centre in Exeter, has factored in over 1,000 different 3-D printing parameters into its AI language to come up with ‘best-in-class settings for more than 100 different components’.
They feature smart extrusion readiness, which analyses the live temperature across four heating zones with rolling stability windows – in essence, providing confidence-scored go/no-go decisions using realistic heating/cooling profiles per material and pre-determined safety thresholds.
Material-aware intelligence also covers extensive variants, including carbon fibre composites and high-temperature polymers. When users say for example: “I need to 3-D print a high-temperature mould tool”, the AI accesses complete profiles taking into account heat deflection temperatures, mould shrinkage, thermal expansion, and mechanical properties for optimal configuration.
Mr Jewell continued: “There is lots more. Task completion orchestrator operates three-tier intent processing; fast pattern matching for simple commands; AI interpretation accesses comprehensive material databases for complex requests; and advanced contextual understanding ensures accurate execution across all scenarios.
“Finally, the AI temperature command handler includes ‘smart chained’ intentions that combine multiple actions. For example, natural requests trigger automated sequences selecting appropriate materials, configuring multi-zone temperatures, evaluating post-processing requirements, and optimising based on embedded engineering data for the specific application.
“It is feature-rich, has a modern user interface, is completely secure and, importantly, has been built for universal integration and agent flexibility, meaning it can be shipped with either an OpenAI-powered agent or a locally-hosted AI agent for online/air-gapped use.”
He concluded: “There is a massive pre-loaded knowledge base, covering all our products, maintenance guides, troubleshooting and practical print flows. We believe we have come up with a ‘true’ AM knowledge bible.”