91ÊÓƵÍøÕ¾ÎÛ

While it’s almost impossible to deny that real-time machine monitoring poses great benefits to machine shops, it’s equally impossible to deny that how much it helps a shop depends on the shop in question. As MachineMetrics Chief Product Officer Rutherford Wilson notes, shops can immediately start using these systems to get information about how machines are being utilized. But, he continues, as they begin to move past clear, simple use cases into more involved work like Kaizen projects, contextualizing the data becomes far more important and requires far more effort.

As a result, Wilson says, many companies best known for machine monitoring are likely to simplify contextualization by moving toward introducing features similar to those available in manufacturing execution system software (MES software). MachineMetrics has recently made strides in this area with Production Schedule Intelligence and other updates to its company’s platform, and has indeed rebranded itself as an MES company. Wilson says the company’s new features and modules add MES-like capability while assisting shop managers in detecting and thwarting production anomalies in real time to prevent problems from throwing production off course.

What Makes an MES?

The direction behind many of MachineMetrics’ recent updates stems from an assessment the company performed around the start of 2023. As part of this assessment, its staff looked into actual use cases for MES software, its value, customer struggles with the software and points where this type of software aligns with the company’s strengths.

Chief among these findings was the realization that shops often needed to perform tasks that seemed basic, but which were interrelated to multiple factors across the shop floor and which gained complexity as a result. Oftentimes, shops would only have information from the machines or the ERP, but the two would not be automatically connected. In extreme cases, some shops — even those with advanced ERP software or complex MES software — would export data to Excel or manually enter data, creating additional paperwork and more chances for errors to creep into the process.

Ultimately, MachineMetrics sought to bridge this gap with a single system connecting its traditional machine monitoring with an ERP-powered schedule. “Clickable Connectors” would directly transfer data between MachineMetrics and shops’ ERP software, but the company also identified four key areas the system would need to address once the systems were connected: OEE analysis, traditional machine monitoring, automated downtime tracking and work order scheduling management. This system was also to leverage MachineMetrics’ machine condition monitoring and predictive maintenance features.

Wilson says the resulting combination of capabilities should give shop managers more clarity on work flow, helping them understand real-time changes to both setup times and cycle times. This includes the use of AI to detect when cycle time begins to show a pattern of deviating from the expected norms, as well as to theorize about potential causes. Wilson says that a balance of jobs leaning more toward repeat orders than new parts makes this more useful, but that high-mix, low-volume shops can also take advantage of the software’s ability to detect anomalies in tool load.

Catching Tool Anomalies

MachineMetrics Tool Anomaly Detection examines tool load and builds a pattern within three cycles, identifying loads that differ from the established pattern to detect tool breaks or chips before they damage a run of parts. Beyond just tools, Wilson says that the examination of load can also help in discovering problems with the density of hard alloys and other materials. It can also detect issues in centering or machine offsets.

Once the software identifies a production issue, operators can issue feed holds to pause a machine immediately after the cycle where the issue was identified. Wilson says this requires high-frequency data to ensure both that the issue reaches the operator in time and that they can set the machine to react accordingly.

While this feature is primarily directed at maximizing tool life, Wilson says it helped solve a lot of other fundamental production issues that may slip through and cause scrap. It can also support customization, as it did for one customer using Swiss-type machines. The customer’s machines had a very standard tool wear pattern, and instead of looking for immediate signs of wear, the shop wanted to detect trends of anomalies. This way, it could detect when gradual tool wear would pass the tipping point for critical failures where tool wear drastically increased and part quality drastically declined. Identifying this point gave the shop the confidence to extend its tool changes from occurring every 100 parts to every 600 parts, increasing both tool life and spindle uptime.

Schedules Subject To Change

The schedule part of MachineMetrics Production Schedule Intelligence uses ERP data about start dates and cycle times as a baseline, but updates based on real-time happenings on the shop floor. The current speed of the machine combines with historical data about the machine in question and the operator running it, adjusting the expected completion time for the combination — and carrying this forward to calculations on future jobs.

MachineMetrics has combined the predictions with an alert system to notify shop managers and operators when jobs are predicted to come in late. Currently, they then rely on a drag-and-drop interface to shift orders, but Wilson says the company hopes to update the software with AI-powered schedule optimization in the near future.

From Installation to Insights

The overall goal of MachineMetrics (and indeed, of the coming generation of machine monitoring and MES systems) in Wilson’s eyes is to take stock of how operators manage their work and to give them actionable insights into production issues. MachineMetrics’ updates are meant to further this goal through new features and additional optimizations in OEE calculation, analytics and downtime classification.

Wilson says that he particularly wants to make certain that any new innovation the company makes is not novel for novelty’s sake, but that it serves a distinct purpose on the shop floor. He sees this as one facet of the company’s overall push to shorten the “time to value” for its software, getting the highest possible amount of usefulness into users’ hands as quickly as possible.