Then and Now: How AI Could Transform a Manufacturing Engineer’s Role in Belfast and Northern Ireland
ARTICLE 1 IN THE SERIES: THEN, NOW AND AI
Then, Now & AI Series 1, Article 1: 20 Projects, a Paint Line and One Very Busy Manufacturing Engineer
If you're only interested in the RoleMap™ AI results, I won't be offended! 😄 Scroll to the RoleMap™ AI Results section near the end to see what happened when I analysed one of my own Manufacturing Engineering roles, and where AI could genuinely support the work without replacing the engineer.
I have spent more than 25 years improving processes, solving operational problems and helping businesses deliver better results.
During that time, I have worked across manufacturing, aerospace, banking, aviation, oil and gas and other complex environments.
I have also accumulated a rather impressive collection of memories involving overflowing inboxes, manual spreadsheets, handwritten charts, unanswered telephones, project plans that seemed to reproduce overnight and PowerPoint presentations created for every possible audience.
So, I am starting a new series.
In each article, I will revisit one of my previous or more recent roles and ask a practical question:
How would I perform this job today using AI, automation and better digital workflows?
This will not be a series written from theory.
I will be analysing roles I have actually performed, the work I was personally responsible for and the problems I had to solve.
I want to show where AI could genuinely support experienced people, and where human judgement, accountability and practical knowledge must remain firmly in control.
For Article1,This will not be a series written from theory.
I will be analysing roles I have actually performed, the work I was personally responsible for and the problems I had to solve.
I want to show where AI could genuinely support experienced people, and where human judgement, accountability and practical knowledge must remain firmly in control.
For Article 1, I am going back to one of my manufacturing engineering and continuous improvement ro 1, I am going back to one of my manufacturing engineering and continuous improvement roles.
And there was certainly plenty to analyse.
A PAINT LINE, A PENCIL AND APPROXIMATELY 20 PROJECTS
At one stage in my manufacturing career, I was managing approximately 20 engineering, capital investment and continuous improvement projects at the same time.
That sounds quite tidy when written on a CV.
In reality, it meant an email inbox that was permanently out of control, manually created project plans, deadlines checked one by one and the constant suspicion that an important approval was quietly gathering dust in somebody else’s inbox.
There was no AI project assistant reviewing the portfolio overnight and warning me:
“Three supplier quotations are late, two approvals are overdue, four project actions have no owner and one installation is about to become tomorrow morning’s emergency.”
There was just me.
Checking.
Chasing.
Sending another email.
And when the emails remained unanswered, ringing people who appeared to possess a desk telephone but were never actually at their desk.
THE WORK BEHIND EVERY CAPITAL PROJECT
For every capital project, I had to manage the tendering process.
I would invite at least three suppliers onto the manufacturing site, walk them through the production area and explain the problem we were trying to solve.
Sometimes I already had a clear idea of the installation, equipment or process change required.
At other times, I wanted the suppliers to use their expertise and suggest alternatives.
Then came the paperwork.
I had to prepare requests for information, technical requirements, project scopes and supporting documents.
These were sent to Purchasing, where I then had to monitor the approval status and continually follow up.
Once a supplier was selected, I reviewed their proposed designs and challenged anything that would not work in the real manufacturing environment.
That could include changing designs for storage racks, trolleys or material-handling equipment so painted components could be transported without scraping their edges, damaging the finish or creating another quality problem while supposedly solving the first one.
When the work was finished, I inspected what had been delivered, confirmed whether it met the agreed requirements and approved the invoices.
The project did not end when the equipment arrived.
The result still had to work.
It had to be safe.
It had to protect the product.
And it had to deliver the promised benefit.
Before contractors could begin work on site, I also had to prepare the relevant safety documentation.
That meant identifying the hazards, assessing the risks and documenting how those risks would be managed.
Contractors were entering a live manufacturing environment containing machinery, chemicals, moving equipment, production traffic and people trying to keep the operation running.
The safety documents could not simply be copied from the previous project and filed away.
They had to reflect the real task, the real environment and the real consequences of getting it wrong.
AI could help draft and structure this type of documentation today.
But AI should never make the final safety decision.
A competent person must still review the work, challenge unsafe plans and retain responsibility for the approval.
That distinction matters.
AI can support accountability.
It cannot replace it.
THE SIX SIGMA PROJECT THAT INVOLVED A LOT OF PENCILS
Alongside the capital projects, I was running Six Sigma improvement work.
One of my most memorable projects involved a powder-coating paint line.
To understand the process variation, I stood beside the moving line and manually recorded measurements as the coated components passed me.
I carried out gloss testing.
I recorded the results.
I plotted the variation by hand.
I created the statistical charts.
Then I wrote the report and built the PowerPoint presentation explaining what the analysis showed.
We improved paint-line efficiency by approximately 30%, delivering the project on time and within budget.
I am still proud of that result.
But looking back, a huge amount of time was spent simply getting the data into a form that could be analysed and communicated.
Today, digital data capture, automated SPC charts and AI-assisted commentary could reduce much of that effort.
The engineer would still need to confirm that the data made sense.
The operators would still hold vital knowledge about what was happening on the line.
The final root cause and engineering decisions would still belong to people.
But the hours spent manually plotting, formatting and recreating reports could be dramatically reduced.
RUNNING THE LABORATORY AND ORGANISING WEEKEND NOZZLE CLEANING
At the same time, I was running or coordinating laboratory and process-control work.
That included monitoring paint, chemicals and coating performance, organising testing and supporting investigations when results changed.
I also organised nozzle cleaning during weekends and shutdown periods.
That required coordinating the right people, ensuring the work was completed and confirming that the equipment was ready before production restarted.
The paint line did not care that it was Saturday.
Apparently, neither did the nozzles.
ANNUAL TIME STUDIES ON THE PRODUCTION LINE
I also conducted annual time studies.
This involved standing directly on the production line, observing how work was actually performed and recording the time taken for individual tasks and production cycles.
I was looking for waste.
Unnecessary movement.
Waiting.
Searching.
Extra handling.
Poor material presentation.
Bottlenecks.
Repeated actions.
Work that looked perfectly acceptable on paper but made far less sense when observed in reality.
This is why I remain cautious when people suggest that AI can redesign manufacturing work without properly involving the people who perform it.
A spreadsheet cannot always show that an operator must walk six unnecessary metres to collect a component.
A system record may not explain why people regularly wait for a tool.
A process map may not reveal that a rack design causes freshly painted parts to scrape against each other.
AI can analyse observations, calculate times, identify patterns and create reports.
But somebody still needs to stand on the line, watch the work and ask the right questions.
CREATING BILLS OF MATERIALS AND MANUFACTURING ROUTINGS
Another part of the role involved creating and maintaining Bills of Materials and manufacturing routings.
That meant ensuring the correct components, quantities, operations, machines, work centres and standard times were recorded.
When this information is wrong, the effects spread quickly.
Materials can be purchased incorrectly.
Production plans can fail.
Inventory can increase.
Product costs can become inaccurate.
People can be blamed for problems caused by poor system data.
Today, AI and workflow automation could help identify exceptions, compare changes, flag unusual entries and route updates for approval.
However, the underlying data still needs technical ownership.
A system can flag that something looks unusual.
An experienced engineer must decide whether it is actually wrong.
MEETINGS, PRESENTATIONS AND MORE MEETINGS
There were also numerous meetings.
Daily production meetings.
Project reviews.
Supplier meetings.
Purchasing discussions.
Safety meetings.
Quality reviews.
Continuous improvement meetings.
Capital approval meetings.
Maintenance planning meetings.
Leadership updates.
Each meeting needed preparation.
Each meeting created actions.
Each audience wanted different information.
Leadership wanted progress, costs, risks, savings and return on investment.
Operational teams wanted to know what was changing and when.
Technical teams needed the engineering detail.
Purchasing needed quotations, approvals and supplier information.
Finance wanted the numbers.
Much of the source information was the same, but I repeatedly rebuilt it into different reports and PowerPoint presentations.
Today, one controlled source of project information could feed different summaries for different audiences.
AI could create a first draft for leadership, another for operations and another for a technical review.
The engineer would check the content, correct the context and approve what was communicated.
That is not replacing the engineer.
It is stopping the engineer from manually rebuilding the same story five times.
DAILY FIREFIGHTING VERSUS THE WORK I ACTUALLY LOVED
All of this existed alongside daily firefighting.
Production problems still appeared.
Quality issues still needed investigation.
Downtime still interrupted plans.
Suppliers still needed chased.
Approvals still stalled.
And urgent work had a habit of arriving precisely when I had finally set aside time for proactive continuous improvement.
The improvement work was the part I enjoyed most.
It might involve researching better ways to prevent corrosion around chemical dip tanks.
It could mean investigating improved powder-coating technology.
It could mean noticing that the business was paying to dispose of chemical IBC containers and asking whether another company might actually pay to take them.
I found platforms based on the idea that one company’s waste could become another company’s resource.
Something treated as a disposal cost could potentially become a source of income.
That is continuous improvement.
It is curiosity applied to the bottom line.
Why are we doing this?
Why are we paying for that?
Is there a safer method?
Could the process be faster?
Could the quality be better?
Could the waste be reused?
Are we solving the root cause, or have we simply become very efficient at managing the symptom?
AND I WAS COMPLETING A MASTER’S DEGREE IN MANUFACTURING MANAGEMENT
While all of this was happening, I was also completing a Master’s degree.
I attended university on day release one day each week and returned for an additional evening every week.
So, alongside approximately 20 projects, supplier tendering, procurement approvals, contractor safety, laboratory responsibilities, weekend maintenance, Six Sigma work, time studies, BOMs, routings, daily production problems, meetings, reports and presentations, I also had academic assignments and deadlines.
Looking back, I am not entirely sure whether that demonstrated exceptional determination or a complete inability to recognise that my diary was full.
Probably both.
WHAT ROLEMAP™ AI IDENTIFIED
I uploaded a detailed version of this Manufacturing Engineer and Continuous Improvement Engineer role into RoleMap™ AI which I have created at AI For Business Northern Ireland
The analysis identified that the role itself should not be automated.
It should be better supported.
The strongest opportunities included:
• Project and action tracking
• Inbox triage and supplier follow-up
• Meeting summaries and action capture
• Reporting and presentation preparation
• SPC and manufacturing data analysis
• BOM, routing and standard-time exception checks
• Supplier tendering and procurement coordination
• Business-case drafting and benefit tracking
• Laboratory scheduling and process-control reminders
• Contractor documentation support using approved templates
The indicative analysis estimated a potential capacity release of:
16.6 hours per week
Approximately 72 hours per month
Those figures are not guaranteed savings.
They are planning estimates that would need to be validated through process observation, real task timing, stakeholder review and pilot testing.
But even if only part of that time were released, consider where it could be reinvested.
More time on the factory floor.
More time engaging operators.
More time solving root causes.
More time improving quality.
More time protecting product.
More time delivering projects properly.
More time preventing problems rather than repeatedly firefighting them.
More time identifying opportunities that improve safety, productivity, capacity, cost and the bottom line.
If you would like to see the full report , please send me a wee message
THINK ABOUT YOUR OWN ROLE
Think back to one of your previous jobs.
How much of your week was spent doing the work you were employed for—and how much was spent chasing emails, copying information, creating reports and updating trackers?
How many different presentations did you build using almost the same information?
How many meetings created actions that disappeared into notebooks or inboxes?
How much valuable experience was tied up in administration?
Now think about your current role.
Which parts genuinely require your judgement, knowledge, relationships and accountability?
Which parts are repetitive, rules-based, document-heavy or mainly involve moving information from one place to another?
Would you like more time to solve the problems that matter?
Would your managers like you to spend more time improving quality, reducing cost, supporting customers, developing people and delivering results?
Would the business benefit if experienced employees had more time to focus on the right things to impact the bottom line?
That is the real opportunity.
Not replacing people.
Supporting them.
Protecting their knowledge.
Strengthening their decisions.
And redesigning roles so talented people spend less time servicing administration and more time improving the business.
THIS IS ONLY THE FIRST ROLE
This is Article 1 in my Then, Now and AI series.
Over the coming articles, I will continue analysing roles from across my own career.
I will look at what the work involved, what created pressure, which tasks still require human expertise and how I would approach the role differently today using AI and automation.
I want to demonstrate this through firsthand experience, not generic job descriptions or theoretical claims.
Because the most useful AI conversations begin with understanding the real work.
What people actually do.
Where time is lost.
Where judgement matters.
Where risk sits.
And where the business could achieve a measurable improvement.
DISCOVER ROLEMAP™ AI
RoleMap™ AI helps businesses rethink roles, skills and automation opportunities without losing sight of the people doing the work.
Upload a:
• Job description
• Skills or task matrix
• Process document
• Role notes
• List of responsibilities
Receive a detailed RoleMap™ AI report and leadership slide deck covering:
• Tasks that should remain human-led
• Tasks suitable for AI support
• Workflow and automation opportunities
• Indicative time-saving potential
• Human oversight and governance requirements
• Future skills
• Suggested future job titles
• A future AI-enabled job description
• A practical 30, 60 and 90-day transition plan
Buy a one-off RoleMap™ AI role review: ( Done for You )
For a Custom GPT license and unlimited internal role reviews, contact AI for Business Northern Ireland:
https://www.aiforbusinessnorthernireland.co.uk/contact-us/
For a wider expert review of your workflows, processes and automation opportunities, book an AI Opportunity Audit.
Visit:
https://www.aiforbusinessnorthernireland.co.uk/
AI should not replace good people.
It should give them more time to use their experience, improve performance, prevent problems and make a measurable impact on the bottom line.
And perhaps it should also prevent somebody from managing 20 projects, running a laboratory, organising weekend nozzle cleaning and completing a Master’s degree with nothing more advanced than a pencil, an overflowing inbox and an unreasonable level of determination.
RoleMap AI: Rethink Roles, Skills and Automation with AI
Frequently Asked Questions About AI in Manufacturing and RoleMap™ AI
How can AI be used in manufacturing?
AI can support manufacturing by analysing production data, identifying unusual process variation, tracking projects, improving reporting, monitoring actions and reducing repetitive administrative work.
The strongest AI use cases begin with a clear manufacturing problem and measurable targets for safety, quality, productivity, cost or delivery.
Can AI help manufacturing engineers?
Yes. AI can support manufacturing engineers with project planning, email summaries, meeting actions, supplier comparisons, data analysis, report creation, time-study analysis and benefit tracking.
It should support engineering judgement, not replace it.
How can AI reduce administration in manufacturing?
AI can reduce the time spent manually updating project plans, searching emails, writing meeting minutes, preparing reports, rebuilding PowerPoint presentations and checking overdue actions.
This can give manufacturing engineers more time to investigate production problems, work with operators and implement improvements.
Can AI improve manufacturing quality?
AI can analyse inspection results, defect records, laboratory data and process measurements to identify patterns that may be difficult to detect manually.
It can support investigations into recurring defects, process variation and potential root causes. Final decisions should still be verified by experienced engineering and quality professionals.
Can AI help reduce manufacturing downtime?
AI can analyse maintenance history, recurring faults, production interruptions and equipment-performance data.
It can help teams identify patterns, prioritise investigations and recognise early warning signs. Maintenance and engineering teams must still validate the findings before taking action.
Can AI support Lean Manufacturing and Six Sigma?
Yes. AI can support process mapping, data preparation, Pareto analysis, statistical summaries, root cause investigations, improvement reports and action tracking.
Lean and Six Sigma still require people to observe the process, involve employees and verify what is genuinely happening on the factory floor.
Can AI support manufacturing time studies?
AI can help organise and analyse time-study data, compare cycle times, identify variation and highlight possible bottlenecks or line-balancing opportunities.
Manufacturing engineers should still observe the real process and speak with operators before changing production standards.
Can AI help manage manufacturing projects?
AI can monitor project plans, deadlines, risks, budgets, approvals and outstanding actions.
It can summarise emails and meetings, prepare project updates and flag tasks at risk of delay. The project manager or engineer remains accountable for final decisions and delivery.
Can AI check Bills of Materials and manufacturing routings?
AI can help identify missing, duplicated or inconsistent information within Bills of Materials and manufacturing routings.
It may also highlight differences between product data, standard times and documented process steps. Experienced employees must confirm any changes before manufacturing records are updated.
Will AI replace manufacturing engineers?
AI is more likely to change parts of the manufacturing engineer’s role than replace the entire role.
Manufacturing engineers are responsible for practical judgement, safety, design decisions, stakeholder management, problem-solving and implementation. AI is best used to reduce repetitive work and improve access to information.
What should AI not do in manufacturing?
AI should not make final safety decisions, approve engineering designs without review or confirm that equipment is safe without a competent inspection.
It should not replace operators, technicians, engineers or maintenance teams whose experience is essential to understanding the manufacturing process.
What is RoleMap™ AI?
RoleMap™ AI is a role-analysis service that examines how AI could support, automate or change tasks within a job.
It identifies which responsibilities should remain human-led, which tasks could be supported by AI and where automation may save time.
What information can be uploaded to RoleMap™ AI?
A business can provide a:
- Job description
- Skills matrix
- Task list
- Process document
- Set of role notes
- Description of current responsibilities
Detailed and realistic information helps produce a more useful RoleMap™ AI analysis.
What does a RoleMap™ AI report include?
The RoleMap™ AI service provides a detailed report and leadership slide deck covering:
- Human-critical responsibilities
- AI-assisted task opportunities
- Automation opportunities
- Estimated time savings
- Future skills requirements
- Role redesign recommendations
- Practical implementation priorities
- A future AI-enabled job description
Is RoleMap™ AI only for manufacturing roles?
No. RoleMap™ AI can be used for manufacturing, engineering, quality, health and safety, operations, HR, finance, customer service, administration and leadership roles.
It can be applied to almost any role where a clear description of the work is available.
How much time could RoleMap™ AI identify?
The potential saving depends on the role, workload and volume of repetitive or administrative tasks.
In this real manufacturing engineering example, RoleMap™ AI identified an estimated opportunity to save up to 16.6 hours per week.
This estimate should be validated through a controlled pilot before wider implementation.
Does RoleMap™ AI recommend replacing employees?
No. RoleMap™ AI is designed to identify how AI can support people, reduce unnecessary administration and strengthen how work is completed.
It also highlights the responsibilities that require human judgement, experience, accountability, communication and practical knowledge.
How should a manufacturing business start using AI?
Start with a role, process or recurring operational problem, not with the technology.
Identify where time is being lost, where decisions are delayed, where information is repeatedly recreated and where better analysis could improve safety, quality, productivity, delivery or cost.
Then choose a controlled pilot with clear measures and human oversight.
Where can Belfast and Northern Ireland manufacturers get practical AI support?
AI for Business Northern Ireland supports businesses with AI automation, AI training, AI strategy, role analysis, AI governance and identifying practical opportunities to improve business results.
Learn more about RoleMap™ AI:
https://www.aiforbusinessnorthernireland.co.uk/shop/RoleMap-AI-Rethink-Roles-Skills-and-Automation-with-AI-p844342894/
Visit AI for Business Northern Ireland:
https://www.aiforbusinessnorthernireland.co.uk/
#RoleMapAI #ManufacturingEngineering #ContinuousImprovement #LeanManufacturing #SixSigma #AIForManufacturing #ManufacturingAI #OperationalExcellence #BusinessProcessImprovement #AIForBusiness #AutomationStrategy #FutureOfWork #BelfastBusiness #NorthernIrelandBusiness #AIForBusinessNorthernIreland