A lot of the Industry 4.0 hype centers around AI and Machine Learning, often with the equally buzzwordy Industrial Internet of Things (IIoT) thrown in for good measure.

Take a look at some of the platform leaders in the space, many offering business intelligence platforms that aren’t purpose-built for manufacturing. They paint a vision of a fully-automated, fully-connected plant floor, to which their enterprise platform can easily pull data and instantly provide insight. Fat chance.

The Most Overused Manufacturing Buzzwords of 2018

Terms like Industrial IoT and Industry 4.0 have come to dominate the industry press and fuel conversations in the manufacturing world. Like the blanket term “innovation”, these buzzwords are often co-opted by marketers to provide a useful shorthand around technological “must-haves”. They are powerful because once referenced, they provide a convenient umbrella under which marketers and industry insiders talk about current trends and propose future states.

But these terms also confuse many (not just the luddites) by placing too much emphasis on the promise of the technology rather than its real-world application.

Take for instance, business intelligence platform leader IBM’s latest take on the Industry 4.0 in a recent marketing campaign. It takes you on a tour of a Model Factory.

In this fancy animation, factory workers (and robots) make shoes in two colors. The KPIs that IBM Watson is allegedly paying attention to are Throughput, On-Time and Risk.

Screenshot from IBM’s Website https://www.ibm.com/industries/manufacturing/industry-4.0-model-factory/

In one scenario, the business intelligence provided is supply-chain related, signalling you that a storm might affect your supply chain, Watson then calculates and mitigates that risk. In another scenario, a bit more realistic, Watson provides you with predictive and preventative maintenance alerts helping to keep one line open so your productivity doesn’t take too much of a hit.  In another scenario, a bit more realistic, Watson provides you with predictive and preventative maintenance alerts helping to keep your production from going off-line.

Real-World Manufacturing is Messy and Complex

In both cases in which Watson saves the day, Industry 4.0 is represented as a all-in-one solution within an already hyper-connected, fully-automated, AI-driven world, in which data appears almost magically and is conveniently abstracted into simple but ultimately meaningless KPIs. The hard work for Watson was already done pulling all that data and making some contextual analytics out of it.

For Most Manufacturers, Machine Learning and AI-driven Business Intelligence is a Moonshot.

The truth is, that in both cases, i.e. supply and value chain management and machine and manufacturing analytics, the actual problems that Watson wants to solve are very complex and nuanced, and will require so much data and technology, that to provide any meaningful insights from an AI-powered platform, a manufacturer may need to invest hundreds of thousands of dollars in retrofits, sensors and connectivity to get even close to a baseline ROI.

Smarter Machines or Smarter Analytics?

What’s missing in a more realistic vision of Industry 4.0 isn’t smarter machines, it’s smarter analytics from the machines you already have.

A few other platforms promise technology breakthroughs through advances in IIoT, in which machines in a production line, raise their hands to self-identify when they require maintenance or aren’t working at peak performance. Some machines even provide a detailed diagnostic to show the maintenance crew exactly what machine part might need replacing.

The concept of a fully realized “digital twin” is also problematic for most manufacturers. While this seems like a natural extension of fully-connected machines, the models are so complex as to be out of reach of all but the most advanced manufacturers, and even then, probably realistic only for their newest factories.

These scenarios conveniently avoid the real-world of the modern manufacturing plant. In most plants, there’s a mix of machines, new and old, some connected and not connected, some finely programmable or still others largely manually operated. We ask: What good is a fully integrated IIoT platform, if the weak link in the chain that leads to unplanned downtime wasn’t the shiny new CNC machine, but something older and less self-aware?

What’s missing in a more realistic vision of Industry 4.0 isn’t smarter machines, it’s smarter analytics from the machines you already have.

The Foundation of Any Industry 4.0 Efforts Will Be Solid Analytics

Many of the value propositions for these new platforms and the ideal factories of the future they promote depend erroneously on AI or Machine Learning or on Hyper-Connected Smart IIoT devices.

The truth is that most productivity gains in the next decade won’t be through Industry 4.0 moonshots like machine learning and AI, or Digital Twinning, but are much more likely to come from run-of-the-mill continuous improvement, applied LEAN manufacturing techniques and good old fashion automation that increases capacity and throughput.

These efforts won’t be informed by Watson, at least not in the short-term, but rather your rank-and-file employees. It will be your line operators, quality assurance, plant and operations managers, and your operation specialists that will make the decisions with help or hurt your bottom-line productivity or make these efforts a success or failure.

By investing in manufacturing analytics that provides context into your production lines, you can achieve a lot of the promised value of Industry 4.0 and IIoT, by providing the one thing that all the model smart factories will require to run at an ideal state: Actionable manufacturing analytics from your plant floor delivered to your employees.

As a MindSphere platform partner, we built SensrTrx to fill that need, providing the manufacturing analytics that can be used for daily monitoring and optimization as well as advanced and real-time modeling. One of the things we liked about Siemens’ MindSphere was an open IoT Operating System approach, in which best-fit software can play nicely together across a lot of other systems you may or may not have.

The Good News? Some of the Hype is justified

Here’s where the hype of smart manufacturing, Industry 4.0 and IIoT might just live up to itself:

  • Sensor technology and PLCs producing useful machine data are ubiquitous and came standard within most machines from the past twenty or so years.
  • Retrofits, once requiring new hardware and a new PLC, can provide meaningful data points through relatively inexpensive peel-and-stick solutions or non-invasive updates to a machine’s PLC.
  • Cloud-based applications (like SensrTrx) can now provide enterprise manufacturing analytics for a fraction of the cost, substantially lowering the barrier to entry for most manufacturers.

All this means that most manufacturers can now afford to benefit from the technology that is helping to create an analytics revolution. If you’d like to get past the Industry 4.0 and IIoT hype and start making evidence-based decisions concerning your plant’s availability, performance and quality measurements today, let’s talk. Contact us for a Demo today.