The Center for Data Innovation recently spoke with Jason Bowes, Chief Operating Officer of CloudNC, a UK-based company developing an AI-powered software to automate manufacturing processes. Bowes explained how CloudNC’s software helps manufacturers program computer-controlled machines, test machining strategies, and produce parts more efficiently.
David Kertai: What problem in modern manufacturing is CloudNC solving?
Jason Bowes: Manufacturing facilities often use highly automated machines and robotic systems, but programming computer numerical control (CNC) machines, which use computer instructions to cut and shape materials, is still largely a manual process. CNC programmers need deep expertise to determine how a machine should make each part, and developing those instructions can take hours or even days.
CloudNC aims to address this through our AI-powered software, CAM Assist, which handles many of the repetitive and time-consuming parts of CNC programming. CAM Assist uses computational geometry, the mathematical study of shapes and their properties, to generate millions of possible ways to cut different regions of a part.
It then combines those options with models of the machine, cutting tools, and physical cutting process to determine which tools to use, which directions to approach from, and how quickly to cut. The software produces a machining strategy, meaning a plan for how to manufacture the part, in seconds or minutes for a programmer to review and refine.
Kertai: What does a machinist see after CAM Assist generates a strategy?
Bowes: The machinist sees a proposed machining strategy alongside a 3D model of the part and its setup, including the starting block of material and the fixtures, which are devices that hold the material securely in place during machining. The results appear in stages so the programmer can follow how CAM Assist developed each step.
The programmer can inspect, adjust, or edit any operation, change the order of steps, or remove results they do not want to use. This keeps the programmer in control of the critical decisions. Once approved, the strategy moves into the customer’s preferred computer-aided manufacturing software, which helps programmers turn the strategy into detailed instructions for the machine. The user can then generate and verify the toolpaths, the specific routes each cutting tool will follow through the material. Every toolpath remains visible and editable, so the machinist can understand what the machine will do before production begins.
Kertai: How do you test whether your software’s recommendations will work on a real machine?
Bowes: Programmers review the generated toolpaths in their CAM software before anything reaches the machine. They can inspect every detail, simulate the machining process, and make adjustments. If something looks wrong or unclear, they can change it.
We also use our factory in Chelmsford, UK, as a real-world testing environment. It produces commercial components, which exposes the software to tight tolerances, meaning dimensions that must stay within very small limits, short lead times, and real production pressures. A dedicated machining cell lets us test CAM Assist while controlling other variables. Data from these experiments helps us validate and refine the mathematical models that drive our software.
Kertai: What happens when your software encounters a part or machining problem it has not seen before?
Bowes: CAM Assist is designed to determine how to machine parts it has not encountered before by combining computational geometry, physical modeling, and machining rules. But machining is complex, and the software has limits. If our software cannot complete a component, it can still produce a partial result that an expert can finish manually within the same CAM system. That can save significant time while allowing programmers to focus their expertise where it matters most.
Our testing in the factory also helps us identify situations where the software needs to improve. As we expand the range of scenarios our models can handle, CAM Assist can support more components and produce more effective machining strategies.
Kertai: Where do you see CloudNC’s technology expanding next in the manufacturing space?
Bowes: We see two frontiers. First, we want to improve CAM Assist by supporting more parts and generating more efficient toolpaths. Second, we want to assist with more of the production life cycle, including quoting new work through tools such as Quote Agent, which helps machine shops prepare quotes more quickly. Ultimately, we want to make manufacturing more efficient and reliable by reducing manual steps and removing bottlenecks throughout the production process.
