Traceability: the least sexy aspect of Industry 4.0

Although the real estate phrase “good bones” is starting to take root in the modern vernacular, how many people actually care about a house’s framework? Let’s be honest, if the décor, design and location are good, a house will sell.

Traceability is like that. It’s the framework that holds the digital manufacturing house together. Important, but not exciting. 

In the food and pharma industries, it’s widely seen as a necessity. There’s a trifecta of factors in those industries to force traceability: regulation, the abundance of counterfeits, and the criticality of recalls. 

But other industries are lagging. Why is that? Like a lot of unsexy things, traceability doesn’t come easily. It takes work. It takes investment. And investment demands return. So what is that return? What value do you get for investing in traceability?

Answering this question starts with an appreciation of what traceability is. Simply put traceability in manufacturing means knowing everything done to every part – every transaction, every movement, every component used – a part’s complete genealogy.

Next is understanding the value of a link between that genealogy and all the data we’re collecting. As we know, our lines are generating boatloads of data every day and putting that data to use is in the forefront of everyone’s mind. This is the heart and soul of Industry 4.0. Machine measurements, machine settings, sensor measurements, measurements made by operators, measurements made on operators; we’re all focused on collecting, analyzing, graphing all the data we can. And this is good, very good.

But isn’t all this data more valuable when linked to product genealogy? Knowing when a machine is out of control is beautiful. But isn’t it even more beautiful to know exactly which parts were processed on the machine during that time?

And isn’t it sometimes the case that problems trace to BOM components? When a printed circuit board fails, isn’t it sometimes due to a bad resistor? Failure shows up in the data from a test operation or an inspection operation but its root cause is a component. In those cases, we need that link: the link between data and genealogy. We need traceability.

Can we know what return we get for implementing traceability?

Can we wrap a number around that? You can start with the most immediate thing: estimating engineering hours saved. Survey your process engineers, product engineers, manufacturing engineers and ask for an assessment of how much time they spend manually making this link: joining production data with genealogy. For some that will be between 5 and 15% of their time. There’s much much more return to be gained by traceability, but this factor alone might just pay for it in 1 to 2 years.

Its unlikely traceability will ever be glamourous enough to become the next buzzword. It’s not realistic to expect everyone to get pumped up about framework. That’s OK. We don’t need to be talking about it all the time, as long as we know not to build a house without it.