Manufacturing Renaissance: Industrial Conglomerate Goes Smart
A Nigerian industrial conglomerate operating bottling, plastics, and packaging plants across Lagos, Ogun, and Rivers states was losing an estimated ₦890M annually to unplanned downtime, energy waste, and production inefficiencies that nobody could quantify because everything ran on paper logs and operator memory.
Overview
A Nigerian industrial conglomerate operating bottling, plastics, and packaging plants across Lagos, Ogun, and Rivers states was losing an estimated ₦890M annually to unplanned downtime, energy waste, and production inefficiencies that nobody could quantify because everything ran on paper logs and operator memory.
The Challenge
The conglomerate had grown through acquisition over 30 years, inheriting three fundamentally different operational cultures. The bottling plant in Agbara ran on a 15-year-old SCADA system that only two engineers understood. The plastics factory in Ikeja used paper-based production logs — shift supervisors handwrote output numbers that were typed into Excel by an admin assistant two days later. The packaging plant in Port Harcourt had a basic PLC setup but no data connectivity to head office.
Unplanned downtime was the silent killer. Equipment failures were only detected when operators heard “unusual sounds” or when a production line simply stopped. Maintenance was purely reactive — nothing was fixed until it broke. The conglomerate estimated (though they had no data to confirm) that unplanned downtime cost approximately ₦890M per year across all three plants in lost production, emergency repair premiums, and overtime labor.
Energy costs were the second major drain. The plants ran generators 18+ hours daily due to unreliable grid power, but nobody could tell which production lines consumed the most energy, when peak consumption occurred, or whether machines were running efficiently. The monthly diesel bill across all three plants exceeded ₦120M, and management suspected at least 20% was waste — but had no data to prove it.
The CEO’s frustration was simple: “I own three factories and I cannot tell you right now, at this moment, how many units we produced today, what our yield rate is, or which machines are running. I find out three days later from a spreadsheet that may or may not be accurate.”
Our Solution
We designed and executed a 24-month Industrial IoT and data integration program across all three plants. The approach was pragmatic — we did not rip and replace existing systems but instead built a digital layer on top of them.
Phase 1 was connectivity. We installed vibration, temperature, and current sensors on 200+ critical assets — compressors, motors, conveyors, injection molders, and generators. Each plant got an edge computing gateway that collected sensor data locally, processed it for anomalies, and forwarded aggregated data to a central cloud platform. This design was critical: Nigerian internet connectivity is unreliable, so edge processing ensured that local monitoring and alerts worked even when the WAN link was down.
Phase 2 was the data platform. We built a custom data lake that ingested sensor telemetry, production counts (from newly installed line counters), energy meter readings, and quality inspection results. Legacy data from the Agbara SCADA system was integrated via OPC-UA adapters. Port Harcourt’s PLC data was pulled via Modbus TCP. Ikeja’s paper logs were replaced with tablet-based digital forms that operators filled out on rugged Android devices.
Phase 3 was intelligence. We deployed machine learning models trained on 6 months of collected data to predict equipment failures 48–72 hours in advance. Vibration signature analysis on the bottling line’s compressors alone prevented 4 failures in the first quarter — each of which would have caused 8–12 hours of downtime. Energy monitoring dashboards identified that 3 injection molding machines in Ikeja were consuming 34% more power than their rated capacity due to worn bearings — a finding that paid for the entire IoT deployment in energy savings.
We also built a custom manufacturing execution dashboard that gave the CEO exactly what he asked for: real-time production output, yield rates, OEE (Overall Equipment Effectiveness), and machine status across all three plants — accessible from his phone.
The Results
Unplanned downtime dropped 45% in the first year. The predictive maintenance system caught 23 potential failures before they caused production stops — estimated savings of ₦340M in avoided downtime costs. Energy consumption fell 22% through a combination of identifying wasteful equipment, optimizing production scheduling to reduce generator run-time, and fixing machines that were drawing excess power.
The CEO now reviews plant performance daily on his phone. Monthly management meetings shifted from arguing about whose production numbers were correct to discussing how to optimize further. The quality team identified a correlation between ambient temperature spikes and defect rates in the plastics plant — leading to a simple HVAC adjustment that reduced reject rates by 15%.
Total ROI was achieved in 14 months. The conglomerate is now planning Phase 4: AI-powered production scheduling that optimizes across all three plants based on order backlog, energy costs, and equipment availability.