How to Boost OEE with Automated Machine Tool Monitoring

Originally titled 'Make Monitoring Successful'

Monitoring enables shop personnel to see the status of a machine and look for ways to increase uptime.

For any machine shop, one of the biggest challenges is knowing what is happening on the shop floor so that activities can be measured, controlled and improved. Automated monitoring systems, such as those based on the MTConnect standard, make it possible to collect data that give managers, engineers and operators real-time information about shopfloor activities to help them make quicker and better decisions.

TechSolve, a manufacturing consultancy firm that offers automated shopfloor monitoring systems, recommends that its clients follow these five major steps to implementing basic yet expandable shopfloor monitoring.

  1. Set objectives.

  2. Review IT infrastructure capabilities and establish system requirements.

  3. Create a data collection plan.

  4. Develop training and dashboard criteria.

  5. Go live when the installation is ready.

Let’s look at these steps in a little more detail.

1. The main objective of monitoring is to get a handle on current performance. A common metric is overall equipment effectiveness (OEE), derived by multiplying machine availability, performance and quality, where each of these factors represents a percentage of an ideal total or optimal level. Availability has typically been the hardest factor to determine. However, it is the one factor that an automated monitoring system can be especially good at providing. The key is the objective data that automated monitoring provides—showing when machines are running or not running, with indications of the reason for any stoppage. A monitoring project plan should focus on availability and how improving it can boost OEE.

2. A review of the IT infrastructure looks at providing Ethernet connections to each piece of equipment, checking the server requirements necessary to run system software components, and determining if a local or cloud-based installation is required. For this reason, bringing your IT manager into the discussion before you begin shopping for a system is highly recommended.

3. A data collection plan must have two parts: determining what data items support the desired metrics (especially those related to equipment stoppages) and determining how those data items can be made available. This requires technical understanding of how to obtain the required data from each asset, and includes assessing the costs and risks of upgrading equipment to obtain that data. Success here depends on input from production managers, maintenance managers (who know the equipment) and IT managers. Close engagement with your chosen monitoring system provider is also essential.

Modern CNCs are likely to need a software-only interface to access the desired data. Older CNCs (with no Ethernet) may require an add-on data-acquisition unit that provides a link to the CNC’s I/O signals. For non-CNC equipment and auxiliary sensors, off-the-shelf devices can be installed to capture digital and analog input regarding motor amperage draw, temperature, vibration, on/off settings and so on. Manually entered data (input from the operator) may also be included in the plan, depending on the requirements.

4. Training at all levels should focus as much on how to benefit from a monitoring system as on how to use or interact with the system. Monitoring can and should benefit everyone, but the gains must be shared and understood by all. For example, wall-mounted data displays often include dashboards because, like the gages in a car, they summarize the trends and show how positive actions move the needles in the right direction. Everyone should understand their importance and role in achieving performance objectives.

5. Once a monitoring system is in place, “go live” when enough data has been collected to be meaningful. Establish a baseline so that awareness of operating conditions can have a starting point once data is on display. This enables positive changes in behavior to register as improvements. At first, make only the most important data public. Encourage openness and feedback. Be ready to tweak the system. Expect some pushback, but keep discussions positive. Finally, remember that there is no return on investment from monitoring. The value isn’t in the data collected, but in the actions taken as a result.