Manufacturing IT: Stuck in a rut?

In recent decades, it seems to me that the manufacturing industry has been a slow — and often times reluctant — adopter of information technology. CAD/CAM and ERP systems have been a big part of the manufacturing world for a while now, but in the majority of plants I visit, this is about the extent of it. Meanwhile, there’s a new revolution happening — M2M (machine-to-machine) and the Internet of Things — that, so far, seems to be ignoring the plant floor.

Technology is becoming a high-value, low-cost commodity. Small, low-cost devices are available off the shelf that can do some extraordinary things. Take the technology in a typical cell phone, for instance — high-definition camera, GPS, enough memory to store a lifetime’s worth of pictures and favourite music, plus the ability to communicate to anywhere, at any time. From your phone, you can open your garage door, turn up the heat in the cottage, check the whereabouts of your children on their way to school, and do your banking. I like to think of this as “horizontal” information technology. We in manufacturing, however, are still stuck in “vertical” information technology. Let me explain.

Our overall manufacturing industry at large is divided into columns, such as fabricating, machining and assembling, and each of these sectors has their own associations and tradeshows. They often operate in isolation from each other, creating their own methodologies, systems, acronyms and even language. Within companies themselves, these same silos exist — stamping, welding, assembly; lead prep, harness assembly, test; cutting, machining, finishing. This is just at the shop floor level. It gets worse. We have shipping and receiving systems, inventory handling systems, production scheduling systems and maintenance monitoring systems. We have engineering departments, production management departments and, trying to stitch these all together, we have IT departments. The data that moves through our plants tends to move up and down through these data silos, with the first point of convergence often being at the executive management level.

There are solutions designed to do it all, but the two biggest challenges are integration and adoption. Manufacturing machines are typically 25 to 40 years old, and the cost of replacing or upgrading is prohibitive, especially in our current cap-ex sensitive world. Discrete and manual manufacturing processes are equally important to have a handle on, but they are often too costly to integrate. Adoption would be a challenge all by itself, but it is amplified by the integration challenge — when machines or processes cannot be connected to the management system directly, they have to be connected by the operator entering the information manually. This often changes the operator’s job, adding additional tasks that are thought of as administrative. Information collected this way is not empirical, and the operator is not as efficient.

When a manufacturer builds a new plant, they build it to last for 25 to 50 years. There will be improvements to the plant over that time, but the basic structure of the plant will stay the same, as will its heavy machinery. The IT components (hardware and software) may be upgraded and replaced, but the basic data architecture will likely remain entrenched. Data will still be collected, moved and managed in silos. Quality departments have their systems, productivity people have their OEE systems, maintenance and tooling people have their own systems, and finance, administration and production management will attempt to assess the company’s financial well-being by looking at the data from their own systems.
Every now and then, it’s healthy to take a “clean white sheet” approach to reviewing and assessing the established and entrenched processes in our manufacturing companies. I think this is especially true when it comes to our IT systems, specifically because of the emergence of M2M technologies. Powerful and flexible devices, at very low costs, are changing the way we move data.

My advice would be to start a small project somewhere in your plant, led by someone who knows that process and your objectives for that process very well. Then pair them with someone who is excited about emerging IT. The best innovation doesn’t come from a top-down driven initiative, big budgets or even corporate culture. It comes from organic experiments like this.

This column originally appeared in the May 2014 issue of Manufacturing AUTOMATION.