A faster path to toolpaths

When it comes to automation capabilities provided by computer-aided manufacturing (CAM) software, the primary focus has historically been rules-based automation (RBA). Developed to capture and apply human knowledge by enabling CAM software to make programming decisions without employee intervention, RBA uses if/ then instructions for programming scenarios specified by manufacturers. While it has helped manufacturers apply best practices, RBA is also an inherently rigid automation tool that is challenging for companies to implement and maintain inhouse, especially as machine-tool complexity continues to rise, new materials become available, and business needs evolve.

To help manufacturers apply institutional knowledge more quickly, easily, and thoroughly, Hexagon has taken the next step in the evolution of CAM programming automation by enabling the use of artificial intelligence (AI) for computer-aided process planning. Hexagon’s ProPlanAI offers new automation technology powered by Nexus, Hexagon’s cloud-based collaboration tool.

The preservation of knowledge was integral to the development of RBA, which can apply programming rules based on material type, surface finish, feature characteristics, and other factors. On top of being time consuming and laborious to implement, however, even slight differences in manufacturing scenarios make it tough to consistently use the automation. With the availability of new machine-learning tools for programming automation, the knowledge of employees who retire from or simply leave the manufacturing workforce can be more easily applied.

Instead of using instructions comprised of conditional statements, ProPlanAI uses existing CAM programmes as the building blocks for ideal processes. To create an RBA rule, a user might specify that if a hole is an inch in diameter and one inch deep, a specific drilling process should be used. If, on the other hand, the hole is half an inch in diameter and two inches deep, then a different drilling process should be used. In either case, a user or a CAM vendor hired by the user is typically responsible for establishing these very specific rules.

With machine learning, manufacturers can use existing programming data collected from previous, validated CAM programmes to apply knowledge from past jobs instead of relying on rules so specific that they may pertain only to a narrow set of scenarios. ProPlanAI can learn institutional practices from a small number of jobs or from hundreds or thousands of past scenarios and can almost instantly identify a validated manufacturing recipe that closely fits the current job.

Putting data to work

The ability to utilise all past programming information is perhaps the single biggest benefit offered by ProPlanAI, which automatically selects strategies that reflect the institutional knowledge and practices of individual manufacturing companies. While highvolume shops that repeatedly produce the same variety of parts can more easily benefit from RBA, job shops don’t see the value in implementing automation that can’t be consistently used. Both high-volume manufacturers and job shops benefit from using programming automation driven by AI because it thrives on the accumulation of data that they’re already producing and requires very little groundwork to implement.

With the click of a button the technology searches through a multitude of past processes to identify the best process for machining any machinable feature. For instance, if a company has preferred methods for machining open and closed pockets, the machine-learning algorithm identifies patterns across hundreds of different data points and automatically suggests a company’s preferred production method. In cases where a company programs a feature for the first time, ProPlanAI can identify previously programmed features that are the closest match to the current job; the user can then update the process and ProPlanAI will store the new information.

ProPlanAI can train and retrain its machine-learning models within a few minutes by searching for similarities and differences among a large volume of data points. For example, a machinable feature could require two or three machining operations, a machining operation may have 50 or 60 different operation parameters, and there could be 30 instances of that feature available in previous projects. ProPlanAI can analyse and cluster all feature and operation information to create models with validated sets of operational parameters that can be used the next time the feature needs to be programmed.

Evolution of CAM

Manufacturers who struggle to fill skilled positions benefit from an increased ability to use data and by empowering less experienced CNC (computer numerical control) programmers with tools that help them succeed with complicated tasks. Because manufacturers use their own pool of data to build programmes, predictions automatically reflect the knowledge and experience of individual businesses, as well as the parts they produce and the machine tools they use.

CNC programmers, who can immediately benefit from the automation provided by AI, also stand to benefit from ongoing development that will ultimately help them accomplish other tasks more easily. On top of streamlining part programming, the technology could potentially help businesses identify latent standard practices, provide notifications about deviations from programming standards, and make better use of product manufacturing information (PMI) for even more automated CNC programming.

While CAM software developers have provided automation tools over the years that help preserve and apply institutional knowledge, the rigidity of traditional automation makes it difficult for manufacturers to remain agile in a continuously evolving industry. With skilled employees in short supply and the complexity of parts and machinery continuing to rise, AI offers higher flexibility, consistency, and efficiency. By capturing and automatically generating solutions that match company preferences and capabilities, programming automation powered by AI saves time, conserves resources such as machine tools and cutting tools, and ensures better results.