When combined with statistical analysis, real-time production and process monitoring can be used to establish parameters for performance, wear, etc. From there, real-time monitoring can be used to track and flag when production cycles and processes measured by machine sensors head outside an acceptable range. Additionally, the data can be used by the MES system to make decisions about production scheduling, ensuring quality and performing preventative maintenance, among other processes.

The actual metrics or key performance indicators (KPIs) that any individual plastics processor tracks will be specific to the company’s business priorities. However, here are a few common measures used to optimize shop floor operations.

Which machines have the best throughput? Monitoring may reveal that, when running the same exact job in Machine No. 1 and Machine No. 2, Machine No. 1 clearly outperforms. With these insights, a plastics manufacturer can decide to schedule a more time-sensitive work order on No. 1 while delegating another job to the slower No. 2.

What is causing downtime? Monitoring and analysis help sort out scheduled breaks, such as 30 minutes of downtime during lunch periods, versus required breaks due to the need for more lubricant.

How close is a machine to requiring maintenance? The ability to monitor changes in vibration, for example, can provide an early indicator of wear that requires predictive maintenance before it creates a downtime issue.

What is increasing the rate of rejected parts? Comparing amperage or pressure curves to historical curves provides indicators into why parts are rejected and what setup changes may need to be made.

What is the rate of material consumption? Monitoring material consumption in real time can help plastics processors keep an accurate, up-to-date record of inventory to facilitate purchasing decisions and production planning.

Is the cavity pressure correct? There is a whole science about capturing cavity pressure and relating it to a machine’s injection pressure and the part recipe. The mold pressure sensor indicates the mold’s overall health (i.e., correct temps, anticipated pressure, flow rate, cavity wear, machine setup, etc.). Capturing this information can help to avoid product defects.

The number of metrics that can be tracked with sensors and real-time production and process monitoring is nearly unlimited. But starting with too many KPIs can ground monitoring and analysis initiatives to a halt — defocusing efforts and consuming valuable staff time in the process. So, the management team needs to first prioritize putting in place a very short list of KPIs most relevant to the business. Once those are established and a plastics manufacturer is regularly tracking metrics against these KPIs, the company can look at incrementally adding new KPIs.

While it’s best to limit KPIs, the growing use of AI means it’s now a best practice to capture and save as much data from all sensors as is reasonably possible. This will help to build a robust historian database with the rich contextual knowledge to fuel AI-driven insights in the future.

Ensuring a successful start

Plastics processors can take other actions to ensure success. Perhaps most important is to start with one machine. We have seen so many companies get excited about monitoring their shop floor, and set up sensors on anywhere from 20 to 40 machines. Then, they come back and say, “Well, this isn't working the way we thought it would.”

The most successful plastics processors start with one machine, get the sensors and monitoring working, and start collecting the sample information that everyone agrees is vital and important to that machine. Then, once the data coming in is valuable to the business, the team can replicate the same approach on the remaining assets.

Machine data considerations

The machine a plastics manufacturer starts with can also determine success. Some teams want to begin with the most difficult machine. That’s a mistake because it is also important to understand the interface and how the software's going to produce the data. For this reason, it’s better to choose a machine where the team understands how the interface works and can be successful. Even if getting up and running is easy, the team will still learn things through the process, and that knowledge will help with the rest of the installation and give everyone confidence that the results will be as expected.

Plastics processors also need to avoid the temptation to start running production or process monitoring on a machine that isn’t performing as well as other machinery since it won’t provide the right answers and usually leads to having to go back and start the process again. Instead, manufacturers should first connect sensors and monitoring to a really good machine from which the team can collect data and create a baseline of understanding. From there, the company now has a springboard against which to measure other machines.

When first monitoring the data from sensors, it is important to remember that sensors placed on different parts of a machine will often produce different results even if the type of sensor is the same. For example, the amperage meter at the incoming power source for a machine will be different from a secondary amperage meter downstream.

Physical implementation factors

There are some physical implementation factors to consider as well. Most successful plastics processors have relied on their own in-house maintenance teams for the install as opposed to outside electricians. That’s because the installers need to be familiar with their equipment, machines and the building.

Additionally, plastics processors should evaluate whether they can rely solely on wireless connections or if Ethernet connections are also required. We’ve seen 5G Advanced wireless networking significantly improve the quality and reliability of communications on the shop floor, but the metal from equipment can affect its reliability.

Also, when doing the electrical set-up, plan for the future by having network drops on every machine. This will ensure that all machines have network capacity as the team expands the installation of sensors across the shop floor and begins to collect data.

Next-generation technologies have lowered the cost and complexity barriers of retrofitting machines with sensors. Together with these best practices, even plastics processors just starting out can quickly gain insights to improve their efficiency and correct issues before they compromise quality. In doing so, they’ll lay the foundation for long-term profitability and growth.