OEE for Botswana's manufacturers — where the losses hide in changeovers, not in big breakdowns.
Botswana's manufacturing base is built on shorter runs and more product changes than a big export plant ever sees — beverage and food lines, packaging, building materials and light assembly serving a domestic market and the import-substitution push. That shape moves the losses: they sit in changeovers, small stops and slow running, not in dramatic breakdowns. Those are exactly the losses a paper log misses. addanode puts live OEE — availability, performance and quality — with reason-coded downtime on your lines, including machines with no digital output.
The losses a small plant pays for daily — and counts never.
The stop nobody logged
A filling or packing line stops for twenty minutes. Someone fixes it, nobody writes it down, and by month-end the shortfall is explained as "a difficult month". Reason-coded downtime turns each stop into a record — duration, cause, line, shift — so the ranked list of what's actually costing you exists before the argument does.
Small stops that never reach a log
The three-minute jam, the sensor wiped clean, the operator waiting for material. Individually invisible, collectively often the biggest single loss on a short-run line — and completely absent from any handwritten record. Automatic capture is the only way they ever get counted.
Changeovers guessed rather than timed
Short runs and frequent SKU changes define manufacturing in Botswana — a domestic market means many products in small volumes. When changeover time is a folklore number rather than a measured distribution, nobody can tell a genuine improvement from a good day.
Four constraints we design around, not apologise for.
- Power events still bite. BPC supply has improved but industrial sites see interruptions and voltage events, and a plant with limited buffer feels each one. Gateways are battery-buffered and log locally, so an interruption enters the OEE record as measured downtime with a cause rather than a gap in the data.
- Mixed-age equipment is the norm. Most lines in Gaborone, Lobatse and Francistown mix new and old machines, many with no digital output. We read them anyway — sensors on the machine (proximity, current, photo-eye) produce genuine counts and stop events without touching the controller, and without a PLC upgrade project.
- Small teams need answers, not dashboards to maintain. A Botswana plant rarely has a dedicated data analyst. Reporting is built to arrive — a shift summary and a Pareto of stops, on WhatsApp or email — rather than to be built by someone in the plant each week. The power side of the same picture lives on our generator and power monitoring page.
- Pula budgets, staged scope. Deployments start with one line — the one where a stop hurts most — and extend once the first line has paid for itself. Cost bands in writing after a remote assessment; no enterprise MES licence to justify before you have any data at all.
How a first project runs: An architecture proven across Southern Africa, hardware pre-configured in Johannesburg, installation by your own electricians or contractors under live remote guidance from engineers in your time zone, and a written scope before anything is ordered. We put a number on the improvement after we've seen your stops — that figure is only worth having when it's yours.
From paper log to live OEE on one line.
1 · Pick the line that hurts
Usually the bottleneck, or the one whose stops ripple furthest. One line is enough to prove or disprove the whole idea.
2 · Count and sense
Read the PLC where one exists; add counting and state sensors where it doesn't. An operator terminal captures the reason for each stop — in the words your team actually uses.
3 · Run one month, then argue
A month of reason-coded data settles what a year of meetings couldn't. The Pareto of stops tells you where maintenance, spares and changeover work should go first.
Runs on the addaNet industrial IoT platform, alongside remote-site monitoring for operations spread across the country. First line typically live in 2–6 weeks.
Botswana production environments — where the points go missing
| Environment | Dominant loss | First thing to instrument |
|---|---|---|
| Beverages and bottling (named client in this sector) | Micro-stops; outage restarts | Counts + run state at the filler; mains presence |
| Beef and food processing | Cold-chain and batch waits | Temperatures + step timestamps |
| Textiles and light manufacturing | Flow imbalance | Pieces per hour per line vs target |
| Mine surface plants | Rare, expensive stops | Vibration and current on the critical list |
| Import-substitution plants | Small lines, no measurement culture yet | One line, four signals, a few weeks |
Where OEE lands first in Botswana manufacturing.
Food & beverage
Dairy, milling, bottling and food plants around Gaborone — throughput per hour and downtime by cause.
Beverage & packaging
Filling and packing lines where changeovers and short runs dominate the loss profile.
Engineering & mine supply
Fabricators and job shops serving the mining sector — machine utilisation and job progress instead of clipboard estimates.
Building materials
Blocks, cement products and steel fabrication — batch counts, plant availability and the cost of each unplanned stop.
Meat & agro-processing
Abattoir and processing lines — rate against design capacity, and where the gap actually opens.
Packaging & light assembly
Where changeovers dominate the loss profile — measured setup times turn SMED work from theory into a tracked number.
OEE and production monitoring in Botswana — common questions
What is OEE, in plain terms?
Overall Equipment Effectiveness multiplies three ratios: availability (was the machine running when it should have been), performance (was it running at rate) and quality (was the output good first time). One number that hides nothing — a plant that looks busy but scores 45% is telling you where more than half its capacity went.
Can we measure OEE on machines with no PLC?
Yes — this is the normal case in Botswana, not the exception. Counting and state sensors mounted on the machine produce real counts and stop events without touching the controller, and an operator terminal captures the reason for each stop. No controller replacement, no vendor software licence for equipment built decades ago.
Is OEE worth it for a small plant?
Often more than for a large one, because a small plant has less slack to absorb a loss and fewer people to investigate it. The deployment is sized to match: one line, the sensors you already have where possible, and reporting that arrives rather than needing to be maintained. If after a remote assessment we think it will not pay back at your scale, we will tell you so.
Our biggest problem is changeovers. Does OEE help with that?
Directly. Changeover time becomes a measured distribution per product pair, not an average anyone can dispute — so you can see which changes are genuinely slow, whether the slowness is setup or first-off quality, and whether a SMED effort actually moved the number. On short-run plants this is usually where the first real capacity is found.
What does production monitoring cost in Botswana?
Entry deployments cover one line and are quoted as a written cost band after a free remote assessment of your equipment list. Starting from the sensors and controllers you already own keeps the hardware component down. We quote in writing before anything is ordered.
Will you promise us a specific OEE improvement?
No. Anyone quoting you a percentage before seeing your stops is selling, not engineering. What we will commit to is that after one month you will have a ranked, evidence-backed list of what is costing you capacity — and that list is what improvement is built from.
Primary sources
- Botswana Daily News (BOPA, gov.bw) — "Manufacturing catalyst for economic diversification" (1 Aug 2023)
- Ministry of Finance (Botswana) — Botswana Economic Transformation Programme (BETP), official page
- Botswana Power Corporation — corporate website home, Load Management FAQ
- Botswana Power Corporation — Electricity Generation Update (2025/26), official web page
Put a number on every stop in your Botswana plant.
Tell an engineer what you run and where the line loses time. We'll tell you which line to measure first — and whether it pays back.