How Delta Beverages cut manual reporting by 2 to 3 hours a day with a real-time operations system

Delta Beverages is one of Canada's leading beverage co-packers. They produce 50,000 cases a day of carbonated, alcoholic, and non-alcoholic drinks for global food and beverage brands, on glass and PET lines that run multiple shifts.

When we first walked their production floor, the operation was, by any reasonable standard, well-run. The equipment was modern. The team was experienced. The processes were documented.

But the data layer hadn't kept up.

Operators logged production counts and downtime events on paper. Supervisors collected those papers at the end of each shift and re-entered the numbers into spreadsheets. A Continuous Improvement engineer named Harman spent 2 to 3 hours every morning calculating OEE, downtime, and throughput from the previous day's records, just to feed the daily management meeting. Preventive maintenance was scheduled on whiteboards. Work orders were tracked on paper that got archived with no visibility into adherence.

None of this was broken. All of it worked. But the whole system ran on a 24-hour delay, and 30% of one engineer's time was being consumed by data entry that machines should have been doing.

That's where we came in.

What we didn't do

The first thing worth saying is what we explicitly chose not to do.

We didn't recommend replacing Delta's ERP. We didn't propose a "digital transformation roadmap" with a Phase 1, Phase 2, and Phase 3. We didn't suggest an off-the-shelf MES system that would have required Delta to retrain their entire operations team on a new interface.

Every one of those approaches is what the big consulting firms would have proposed, and every one of them would have failed for the same reason: they ignore the actual problem.

The actual problem at Delta wasn't a missing system. It was that the systems they had didn't talk to each other and the people on the floor were doing the integration manually, with pens and paper.

The right fix was smaller, more focused, and faster to ship.

What we did instead

We spent the first weeks of the engagement on the floor, not in slides. We watched operators fill in downtime reports during shift changes. We sat with Harman as he typed numbers into spreadsheets. We mapped maintenance flows on the whiteboard with the maintenance manager. By the time we wrote the first line of code, we had a clear picture of every gap and every workaround the team had built to compensate.

Then we built four things that fit together as a single system.

A custom production application that line operators use on tablets, right next to the equipment. Production counts and downtime events get entered as they happen, with the operator selecting categorized downtime reasons from a structured list (instead of writing free text on paper). Data validation runs in real time, so duplicates and obvious errors never get into the system.

A suite of Microsoft Power Apps for maintenance. One app for maintenance leaders to schedule and assign preventive maintenance tasks based on actual asset usage instead of a fixed calendar. One for technicians to receive their work orders, complete them, and attach photos or notes. One for administrators to manage SKU configurations and asset settings.

Ten Power BI dashboards that present real-time and historical data in formats designed for different roles. Real-time hourly throughput for floor supervisors. OEE by line and shift for plant managers. Maintenance adherence and work order completion for the maintenance leadership. Long-term trend analysis for continuous improvement.

A SQL Server backbone that ties everything together. Production data, downtime records, maintenance logs, asset usage history. All structured for fast retrieval into Power BI and built with the expectation that the data would be queried for years.

The whole system was designed to integrate with Delta's existing infrastructure, not replace it. The ERP stayed where it was. The team's workflows changed in small, intentional ways rather than being wholesale reorganized.

What changed

The shift on the floor was immediate.

Operators stopped filling out paper forms. Tablets replaced clipboards. Downtime events that used to surface 24 hours later (after Harman finished his morning calculations) now appeared on dashboards within minutes. The daily management meeting changed character: instead of reviewing what happened yesterday, the team could discuss what was happening right now.

Reactive problem-solving became proactive. When a line slowed down, leadership could see it and respond before the next shift, not the next morning.

Preventive maintenance became predictive. Assets accumulated run-time hours in the system, and PM tasks triggered when usage actually warranted them, instead of when the calendar said so. Maintenance audits, which had previously meant digging through archived paper, became a database query.

Harman's 2 to 3 hours a day disappeared. That time, 30% of his role, redirected from data collection to actual continuous improvement work.

Andrew Elias, Delta's VP of Operations, put it this way:

The results have been outstanding. The project provided us with real-time visibility over performance metrics, significantly improving our ability to monitor and optimize operations. Adherence to maintenance tasks has improved, reducing downtime and enhancing overall efficiency.

The full project paid for itself in under a year, just from recovering Harman's time. Everything else (faster response to floor issues, better-targeted maintenance, cleaner historical data) was upside.

The lesson

There's a pattern in B2B operations that we keep seeing. A company knows it's inefficient. They go shopping for a "digital transformation" partner. They get sold a six-month roadmap, a year of implementation, and a transformation program that promises to revolutionize how the business runs.

Eighteen months later, they have a new system that nobody uses and the operators are still filling in paper forms.

The Delta project worked because we did the opposite. We started small. We built around what was already working. We shipped fast. We trained the people who'd actually use the system. And we left when the work was done.

That's the model.

If your operation is still glued together by spreadsheets, paper logs, and end-of-day reporting cycles, the answer probably isn't a replacement system. It's a focused intervention that closes the data gaps and gives your team real-time visibility, designed around what you already have.

That's what we do.