Industry 4.0: Deploy IoT Now Using a Phased Approach

By Brian Merckling

There’s been a lot of buzz around how IoT will create “Industry 4.0” as we build sensors into everything from refrigerators to forklifts to gas turbines. However, recent studies show that it’s not as easy as it sounds. Substantial challenges exist in separating hype from reality and being able to create a true plan for transformation.

As with most big shifts, change won’t happen all at once. A scaled approach is the best way to embark on the journey to Industry 4.0. Below I share some ideas that you can bring to your customers.

First, avoid boiling the ocean. The way to failure – or certainly a big stalling point – is to present a solution that means changing the whole business at once. That’s going to hit cost and change-management hurdles out the gate. And who wants to wait two, three, or four years for a project to get underway?

In a recent study, McKinsey found that only 30 percent of technology suppliers and 16 percent of manufacturers have an overall Industry 4.0 strategy in place, and just 24 percent have assigned clear responsibilities to implement it. Your customers need a plan and solution they can ingest one bite at a time.

Here are the essential ingredients you, the partner, can provide:

Begin With an Industrial IoT Hub

The essential foundation is an Industrial IoT hub that will scale to fit the needs of the business. The alternative is to spend millions of dollars and years in development to build and maintain a robust back end with industry-specific front-end applications and the ability to monitor thousands of remote assets on the edge of a network. That takes rigorous project planning, capital investment and time — up to three years for a full deployment.

Instead, partner with one of a selection of vendors that offers Industrial IoT hubs. Look for flexibility, scalability and interoperability with the customer’s existing systems and the ability to work with a variety of hardware. Your customer might not be ready to abandon legacy systems, and flexibility will allow them to run in parallel for a time. You have the option to layer in your own implementation, support or app development expertise to complement that.

When looking for a supplier, bring along a checklist of requirements to find an architecture that will bridge today’s needs and the future IoT vision.

Look for:

We already mentioned that it would take a lot of capital to spec and build from scratch. Or, maybe your customer just isn’t sure what needs to be connected. Start with easy-to-attach sensors that work in an industrial environment. This creates a test-and-learn sandbox that speeds figuring out the end goals.

Speaking of beacons, look for ultra-low-power beacons developed using Bluetooth Low Energy, or BLE. They deliver better performance than RFID tags. Industrial-grade BLE beacons work in conjunction with gateways that can locate them, whether stationary or in motion. Such a system can be up and running in days without costly infrastructure or waiting for custom-built equipment.

Benefits of Machine Learning

Machine learning is a method of data analysis that automates analytical model building. Using algorithms that iteratively learn from data, machine learning allows computers to find hidden insights without being explicitly programmed where to look.

The iterative aspect of machine learning is important, because as models are exposed to new data, they are able to independently adapt. They learn from previous computations to produce reliable, repeatable decisions and results.

Machine learning can improve the accuracy of asset location in indoor use and eliminate the issue of “ghosting,” where errors in reading tags, particularly RFID tags, can result in erroneous data, whether the asset is overlooked or incorrectly placed. Ghosting occurs when a gateway can’t interpret the exact location of an asset, especially one moving or in a crowded space, or near water or metal. Gateway design is key in eliminating the appearance and impact of ghost reads.

The End State: Cognitive Computing for Predictive Analytics

To solve challenging use cases that involve advanced analytical analysis, once you have collected a large data set, you could add cognitive computing to help customers interact with that data in an intuitive, natural way, and be able to ask questions. For example, as opposed to just knowing where a wheelchair is, they can answer, “Where should it be? Where does it get misplaced most often? What has it come in contact with? What’s the best place for it to be stored?”

Cognitive computing requires collecting months of data points for the system to learn from, so we recommend that it’s added as a third phase of deployment.

Conclusion

Industry 4.0 can be reality today. Help your customers embrace IoT by looking for suppliers that deliver flexible systems that work well with others. This gives you the option to bring multiple partners to the solution, add more value and avoid getting you or your customer boxed into a single provider’s ecosystem. Starting small will build big long-term revenue.

Bryan Merckling is the CEO of THINaër, an Advantix company.