The importance of connecting your operations

Connecting your legacy systems and machines with the Industrial Internet of Things can open the door to advanced data analytics and actionable metrics, paving the way for process improvements and business objectives.

Most modern machines can be configured to communicate over protocols like MT Connect or OPC. This could be through native functionality, purchased add-ons, or protocol translation products like Kepware. They output a suite of data, such as feed rate, status, number of parts produced, and many more. Legacy machines don’t have the ability to capture or communicate this data, but you can still bring them online using sensors or data feedback devices and an IoT gateway.

Here are 5 steps to a successful digital transformation of your legacy systems with IIoT:

1. Define goals for your IIoT project

Many IIoT projects fail because they lack solid goals rooted in a clear business case.

Therefore, the most important part of any IIoT project is defining clear, measurable goals. These goals should be driven by a solid business case.

Your goals should be measurable by KPIs and pre-defined measures of success. In our experience, good starting places for IIoT are process visibility, operator efficiency, inline quality, though there are countless more. You know your operations best, you and you know where the opportunities for improvement in your legacy systems are. The better you can define a successful outcome, the more likely you’ll be to select a solution that guarantees you’ll get there.

Without a foundation in key business metrics, the new technologies a company implements will fail to live up to their promise.

2. Determine the questions your new data should answer

You should know what questions you want your data to answer before you set out to collect it. Determining the answer to this question should involve a conversation with your production staff and engineers.

Do you want to understand machine performance in a more holistic way? Maybe you’re interested in gathering machine data to support a predictive maintenance initiative. Either way, understanding what kind of data you’d like to collect will determine the departments, machines, and software systems you’d like to target for digital improvement.

Analyze, categorize, and summarize information on legacy systems that can be improved. Chances are you already have an idea of areas for improvement–now you can use data to pinpoint it and make informed decisions.

3. Bring facility infrastructure up to date if necessary.

Not all legacy systems are equally ready to come online. Nevertheless, there are simple things you can do to improve connectivity.

You’ll also need to consider any changes you need to make to your facility or network infrastructure. For example, if you’re working with machines without any connectivity, chances are your facility isn’t set up for Ethernet connectivity. In this case, you’ll need to install Ethernet drops and run cables, which will require approval from facilities.

Keep in mind that any changes that need to be made will take extra time to implement. It’s also crucial to gain the cooperation of those in your company with IT expertise, as you’ll likely need their approval to make any changes.

4. Map out a plan for collecting data from legacy machines using sensors

Every factory’s legacy systems are a little different.

Thus, you will need to compile a list of all the machines in your facility. They should go into 3 categories:

  • Modern machines satisfy two requirements: 1) they can connect to your network, and 2) they speak a protocol your machine monitoring software can read (Tulip can read any protocol).
  • Analog machines do not have network connectivity and do not speak any protocol. These machines will require analog monitoring. These machines will be monitored using external or non-invasive sensors, which will be mapped to machine state.
  • Anything in between: machines in the middle may be able to fulfill both requirements with a software upgrade or a physical network expansion card. This will require further research and consultation with whoever is in charge of your machines as well as the machine’s manufacturer.

The following steps primarily concern analog machines.

Ask yourself:

How can I collect the data I need to meet my goals? What’s the most reliable way to know what I’m trying to identify?

A good example is identifying whether a machine is running. Your instinct might be to track whether or not the conveyor belt is turning–but it’s possible that the conveyor belt is turning without any parts on it. A better indicator, then, of whether the machine is running is whether there are parts coming off the line.

So, a simple and reliable solution to determining whether the machine is running would be to install a break beam sensor where the parts come off the line. As long as the break beam sensor detects motion from the part leaving the conveyor belt, you can map the state of the machine as “running.”

5. Integrate new sensors onto legacy machines.

Once your sensors are in place, you can map sensor outputs to machine states. This may seem like a daunting task, but a no-code software platform like Tulip makes it easy to integrate machine data into powerful apps and analytics.

You can now use this information to monitor metrics such as OEE (overall equipment efficiency) and OPE (overall process efficiency), pinpoint issues and areas for improvement, and inform your processes directly.

Tulip’s platform provides an intuitive, no-code ecosystem to connect and utilize data from legacy machines alongside modern machines. With Tulip, companies can undergo a digital transformation and integrate Industrial IoT technologies with their legacy factory machines. For a low-risk way to experiment with IIoT, try Tulip’s Factory Kit.