The job of a manufacturing plant manager is a challenging one. They are not only responsible for managing the production, efficiency, quality, safety, and overall plant operations on a day-to-day basis. They are also expected to be constantly on the lookout for improvement opportunities to ensure long-term competitiveness of the operation. This mandate requires a delicate balance of focusing on the present with an eye on the future, juggling short-term challenges while enabling long-term opportunities, and managing risk with every decision made.
In the technology climate we find ourselves in, however, it can be increasingly difficult to remain focused on the realistic and sensible, particularly as it applies to the manufacturing industry. The Industrial Internet of Things (IIoT) hype machine, for one, is in overdrive and there is no sign of it abating.
Plant managers, not to mention other IT and operational leaders, are inundated with offers enticing them to invest in new products or services, each promising compelling ROI narratives. Not wanting to be left behind, those with the budgets and the bandwidth are quick to move forward, expecting scalability to naturally occur once early use cases are proven out. And if the new solution is not the magic bullet? Well, at least they tried, and they can use the learnings to decide on the next investment. On the flip side, there are those with limited budgets and bandwidth that find themselves stuck trying to separate the hype from the truly promising opportunities.
Both these ways of thinking can be frustrating for the operators on the plant floors. Either they are forced to constantly learn new tools only for them to be taken away — as if their jobs are not hard enough already — or they never see their companies bringing in new technology to improve efficiencies because management can’t seem to get out of their own heads.
Almost every manufacturer I’ve spoken to is asking this same question. How do we separate the hype from the reality, and identify opportunities that can really address real-life manufacturing concerns?
To examine how manufacturers can act more pragmatically when it comes to implementing new technology, and Industrial IoT solutions specifically, we need look at this from two perspectives — that of the small-to-midsized manufacturer and that of the enterprise manufacturer.
The pragmatic plight of the small-to-midsized manufacturer
Approximately 90% of manufacturing establishments in the US today fall into this category, defined as a company that has less than 500 employees, according to the National Association of Manufacturers. Chances are, they are operating with limited budgets for “experimental” innovation, they don’t have a large IT staff to take ownership of new technology initiatives, and they are working with a somewhat ageing set of machines. These manufactures usually spend less time thinking about how technology will transform their business over the next 10 years, and more time trying to solve today’s problems around downtime, efficiency, quality, and ensuring a smooth operation. They need solutions and they need them now. Their customers are demanding more agility, transparency, and price competitiveness, so these plant managers need to make very pragmatic investments with these factors in mind.
These pragmatic manufacturers need to ask themselves “What critical business problem do I have that I can apply technology to?” instead of “Where can this new cool tech transform my business?” The idea is to look for a solution to an existing problem, as opposed to going fishing for a problem to fit a vendor’s solution.
Doing nothing is certainly not an option
The world of manufacturing is changing far too quickly for the SMB manufacturer to sit idle.
For example, I visited a baby formula powder manufacturer several years ago. Their maintenance program involved engineers with hand-held data collection equipment walking through the facility on six-week circuits. Every day these engineers would manually capture a snapshot of health data from some of the machines — temperature, vibration, lubricant levels and so on — conduct manual analyses, and store the results for future archival. Not only was this program not scalable, the manufacturer was frequently reminded that the time between ‘Potential failure’ and ‘Functional failure’ on the P-F curve — a graphical representation of the health of equipment over time — can often be less than six weeks. While they optimized their routes, it did little to prevent a major failure of the blower fan of their dryer — a single point of failure for the facility — a couple days before my visit. Unfortunately, this wasn’t the first time an incident like this had occurred. And it likely will not be the last given their current process.
This narrative is not unique to this particular manufacturer, and many reading this article will relate to that scenario.
Needless to say, this is an opportunity to take a pragmatic approach to technology decisions on manufacturing floors. Manufacturers need to look for solutions that address the root cause of the problem at hand — enabling real-time data collection, as opposed to six-week snapshots, which delivers trending data to warn of failures before they occur. Machine learning and artificial intelligence in these scenarios may be great, but it may be overkill as a first step.
In our experience in working with manufacturers, the common problems plaguing them include unplanned downtime, diminished machine performance, and substandard quality output. A proper solution will allow them to see immediate dividends so that the ROI becomes more noticeable more quickly.
The enterprise manufacturer’s abundance of data can be stifling
For the enterprise manufacturer — one that employs more than 500 people and oversees multiple plants spread out across the globe — there is still a need to be pragmatic, but not in the same way as their smaller peers.
Larger manufacturers have been collecting data for decades and their need for pragmatism lies in the long-term stability and efficiency of their technology investments. Over the years these manufacturers have invested in a number of IT initiatives to optimize processes, bringing in new technologies, and enabling various parts of their businesses to solve problems independently. And now, the maintenance of these solutions has started to become a complicated nightmare. Enterprise manufacturers are increasingly looking for a single source of truth, centralizing independent ERP, MES, IoT and SCM solutions into one.
These manufacturers are either out there looking for an innovation in enterprise software delivery, or they are investing in developing their own single source of truth, essentially forcing the IT departments at these manufacturers to functionally operate as software companies.
Enterprise manufacturers may theoretically have the resources and manpower to take on this task, but they, like the smaller manufacturers, would rather spend time on their core business. Add to the fact that they might not have the skills to enable contextualization and continual innovation of data models, and you can start to see where enterprise manufacturers might struggle.
Simply put, plant managers in enterprise corporations must not be distracted by endpoint solutions for data capture, and should instead emphasize integration and data contextualization. They need to be asking how their various systems will interact together, and how data will be shared across systems and across their broader, often global, enterprise.
As they separate hype from reality, they need to be asking how these systems will enable real-time, data-informed business decisions from their shop floor to the top floor, with all their users working off a single source of truth.
Anything else is not solving for the long-term scalability and agility that enterprises desperately need, and is merely responding to the hype cycle.
Separating hype from reality
While the solution marketplace continues to get crowded with some very cool tech, manufacturing leaders now more than ever need to remain focused. Our experience with thousands of manufacturers yields the following checklist to separate hype from reality:
- Identify the business problem and the 'need' first. Then go look for a solution. Not the other way around
- You have to collect, store and manage data before you can consume or analyze it. Start from the basics
- Think about the long-term implications of the support and maintenance burden on your team
- Consider long-term aspects of scalability, agility, integrations, and data contextualization
Manufacturers need to consider their technology investments pragmatically, focusing on solving real business problems today while keeping an eye on the long-term implications of their decisions.