In our earlier blog post, we talked about the infrastructure and technical issues that mean manufacturers struggle to integrate and scale digital projects effectively. But there’s another side to this I’d like to explore here: the people and process issues that can hold back digital manufacturing innovation.
Nearly all the manufacturers I speak to tell the same story: there are significant trust issues between their IT and Operational Technology (OT) teams, and this lack of trust is playing a big part in their failure to get digital manufacturing pilots into mainstream production.
Fundamentally, the IT and OT worlds have different priorities. The OT team has often been running the factory for many years, supporting tried and tested systems that have been around for a long time. These systems only tend to remain secure because the factory operates largely in isolation from IT and the internet, effectively making it air-gapped from the outside world. IT then comes into this settled scenario with new ideas that the OT team may see as potential risks to safety and availability - the two things they care about the most. To the OT team, these new ideas can feel like outside initiatives that are being imposed on them, with a lack of understanding of the inherent risks.
One CIO told our CEO how a consultancy firm had come in, captured the customer’s data and handed it to their data scientists, who’d then presented back a finalised AI solution. The factory management team and the OT engineers felt that they hadn’t been consulted and the project was effectively a done deal, using technology that locked the company into a specific solution for years to come. As a result, the OT team didn’t feel any ownership over the project; their feedback was “it just felt like something that was being done to us”, so they weren’t invested in getting it implemented.
I’m a big believer in a teaming, or a co-creation approach to digital manufacturing projects. If you include people in the development process, they’re much more likely to trust the end product and fully deploy it. Transparency is key. If you’re completely open with the OT team, and share the logic behind the intellectual property, they can see the benefits for themselves. My mantra is ‘get the AI out of the hands of the data scientists and into the hands of the engineers’ – and it works.
As an example, we use AI algorithms to help customers save energy. Getting the OT team involved in creating their own model based on the algorithm and using their own data is key. The model then acts as their trusted advisor. The engineers remain in control, but now have a digital coach, as they experiment. As the model proves itself, they still double check things to be absolutely sure, but they begin to trust it to operate some factory functionality independently. These sort of pilots prove to be a success, building trust and eventually freeing up some of the engineering team to focus on new areas.
It’s human nature to resent someone coming in recommending a rip and replace approach. In my experience, most manufacturers have tried a range of initiatives and achieved some success, so it makes sense that they’d be reluctant to start a new pilot if it means throwing that away. I’ve also come across understandable suspicion of technology that ‘wasn’t invented here’.
The answer is to treat digital manufacturing pilots as brown field developments that build on existing work. For example, an organisation may have already deployed predictive maintenance so they can better understand when downtime will occur, but what about then upping the ante with prescriptive guidance? For instance, how can we keep something running until it’s more convenient to fix it, or how do we minimise any down time, like with a F1 pit stop?
Unsurprisingly, many digital manufacturing pilots fail to scale because they can’t demonstrate a rapid Return on Investment (ROI). ROI is even more critical today, as manufacturers are battling the negative effects the pandemic has had on their markets. Margins are tight and a lot of businesses are constrained by costs.
The pilots that do scale in this tough environment are the ones that provide a clear baseline and can quantify the cost savings and show a positive impact on, for example, the Overall Equipment Effectiveness (OEE) score. The key to achieving this is setting the pilot up from the start to be a proof of value exercise. It’s about turning the automatic ‘we can’t afford this’ into ‘we can afford this if we make these proven savings’ – and being able to get a ROI in under eight weeks.
This may also involve lifting the discussion to the C-suite level and getting the chief financial officer to look at the broader business case. If they could improve their OEE score, or reduce the number of maintenance call outs, or improve quality, could the savings be used to subsidise often long overdue infrastructure refreshes?
Working out the ROI can sometimes help justify what might seem to be expensive new investments. For example, a recent 5G pilot identified savings of £30K a week, by using 5G-enabled headsets to allow remote experts to guide employees, cutting the need for travel. So the savings from just one use case already more than covered the deployment costs.
At BT, we believe a co-creation approach is the most effective method for overcoming the people- and process-related barriers to scaling in digital manufacturing. We have a strong track-record of listening to our multinational customers and building a collaborative way of working that delivers results. Our consultants could work with you in your factory space, to explore how you operate and what future options could look like. We then investigate with you each of these options, putting together a proof of value exercise to support your choices and establish the best way to scale the benefits you want to achieve.
If you’d like to find out more about how we can help you bring digital manufacturing to life, please get in touch with your account manager.