Attendance and Safety Records Give Manufacturing Executives an Early Warning on Workforce Problems

Manufacturing plants record who came to work, who didn't, and what went wrong on the line. Attendance figures, unplanned absence rates, near-miss reports, and quality records get reviewed for what they say about output and cost. The same numbers can also indicate whether employees feel safe and respected enough to keep showing up. Executives adding analytics and AI on top of plant data are leveraging records that also reflect how the workforce is being managed, provided the processes producing those records are stable enough to trust.

Paul Mcloughlin, Senior Manufacturing Consultant at Bosch, has worked in manufacturing and on factory floors since his early twenties and now advises companies on change management and lean foundation tools. He works alongside an Industry 4.0 specialist on engagements with manufacturers that are mapping how to pull information from existing equipment and deciding where AI belongs in the resulting workflow. Mcloughlin starts those programs at the value stream, going step by step before anyone chooses what to measure.

"As an executive or part of the leadership team, if you can't see the full picture, you are never going to make the right decision. If you're not getting all the relevant information, you don't know whether you're turning left or right at the junction," says Mcloughlin. Manufacturers generate large volumes of operational data. What reaches the leadership team is often a narrower account, and Mcloughlin's consulting work centers on that gap. He points executives back toward numbers their own operations teams already produce.

Attendance carries the earliest warning

Mcloughlin frames plant performance around five measurement categories that most operations teams already track: quality, safety, cost, delivery, and people. Four of them describe the product and the operation. The fifth describes the workforce building it, and he treats the people measurements as a strong indicator of everything else. Absence and safety records already exist inside operations reporting, without any separate survey. "You look at attendance, you look at casual absenteeism, and from a safety point of view you look at accidents or near misses," says Mcloughlin. "If you've got a really high level of unplanned absenteeism, the consequence of that is it knocks onto your production volumes, it knocks onto your quality, and it can have an impact on safety."

People leave a manufacturing organization at different rates depending on where they work in it. Senior and middle managers tend to stay. Operators and line staff go more often, and replacement costs recur each time a position turns over. Mcloughlin describes what constant recruiting at that level does to the rest of the operation. "At a shop floor level, manufacturing or operational level, people will move and you'll get turnover," notes Mcloughlin. "But that's also where the most valuable capital of your business is, the people who are building the product for you and the service. You're constantly recruiting at that level, and by doing so, quality suffers, absenteeism suffers, because people start to lose interest and become disengaged."

Dashboards give a partial account of a site. Mcloughlin walks the workplace during assessments and looks at the work itself. He checks how tools and materials are organized, whether people are struggling to finish the work in the time allocated, and whether standards exist both for doing the job and for escalating a problem. He is also careful about assigning every operational failure to culture. "You can have bad quality, and bad quality might not be driven by a toxic environment. It might be driven by a lack of process or a lack of training," says Mcloughlin. "For me, safety and attendance at work are really key, because people won't come to work if they don't feel safe, and they won't come to work if they're not being treated with respect."

Clean data comes before AI

Mcloughlin works with companies building out a digitalization and AI strategy, and the order of the work is where he starts. He maps the value stream first, identifies which steps add value, removes waste from the ones that do not, and only then decides which points in the process are worth instrumenting. His concern is with companies that run those steps in the opposite order. "A lot of the problem is they don't understand that if you don't have clean, solid foundations to your business processes, then what you're doing is just putting poor information into a system," says Mcloughlin. "You're capturing poor information, and then you're looking at trying to get AI to make decisions off very poor or clustered information."

Mcloughlin describes AI taking over the physical checks and the data crunching a person would otherwise do by hand, which frees that time for the parts of the job that add value. Authority over the outcome is still a non-negotiable. "AI is secondary to cleaning your data. Have your clean data in place, then you look at how you overlay AI to help decision making," explains Mcloughlin. "Key to that is never take the person out of the process, because AI won't make the decisions. AI will give you options. The person will make the decisions."

Announcements that pair AI adoption with headcount reduction cause a separate problem. Mcloughlin points to public statements from large IT companies, insurance companies, and banks that name a number of roles being cut without describing what the technology will do, which in some cases may be base-level tasks. Employees who aren't told the specifics make their own assumptions. "Leadership has to have a change management plan. They have to have a business strategy," Mcloughlin says. "If they want to implement digitalization and AI, they have to communicate to the people, bad news or good news."

Clearing the path for escalation

After an assessment, Mcloughlin identifies relevant data and gets teams together to talk through what it shows. Building enough trust for those conversations to be worth having is the part he describes as difficult, and he places the responsibility for it with leadership. A working escalation route is the next thing he asks about. "From a leadership point of view, have the confidence that you've got the right team around you," says Mcloughlin. "As things happen in the organization, how does it get escalated? Is it clearly escalated? Is there a plan to resolve the issues? Some of the smallest issues in any work environment can grow and magnify over time."

Mcloughlin runs workshops that put a production team around a table for structured problem-solving. Getting what that room knows in front of the leadership team requires no new technology, and it's available before any model is deployed. He puts a number on what those rooms contain. "You'll have 10 people sitting around a table with 10 years worth of experience each, and you go back to the company and say you've actually got 100 years worth of experience sitting around this table," concludes Mcloughlin. "The people who turn up every day to do the work know what makes a difference."