In late 2022, I was handed a project that, on paper, looked like a slam dunk. A 50,000-square-foot distribution center needed a lighting retrofit. The existing T5H0 fixtures were failing. The warehouse manager wanted energy savings. My boss wanted a modern, ‘connected’ facility.
I’d spent three years handling print procurement. Wiring diagrams, DALI drivers, IoT platforms—that was new territory. But I’d read the case studies. ABB’s IoT platform connect technology could slash energy use by 60%. The ROI projection was eighteen months. Easy, right?
What followed was nine months of cargo-cult decision-making, a $3,200 redo bill, and one very uncomfortable meeting with the VP of Operations. Here’s what I learned. Not from a sales brochure—from the mistakes I personally funded.
The ‘Smart’ Part Is Not the Light Bulb
My first mistake was thinking the hardware was the hard part. I spent weeks comparing luminaire specs, but I filed the “ABB smart sensor” and “ABB Ability platform” paragraph under future consideration. My thinking: we can swap the fixtures now and add the smart stuff later.
That was wrong.
Here’s what I didn’t understand: the sensors, the gateways, the control bus—that’s the retrofit. The LED tubes are a commodity. The intelligence is in the network that connects them. If you run the wiring without planning for that network, you’re not running the right kind of wire. (Or you’re running it to the wrong locations, which is what I did.)
“We designed the sensor grid to fit the 20-foot ceiling layout. The contractor had already run the conduit for lighting only. The sensors needed Low Voltage access—which wasn’t there.” — paraphrase of our site survey report, March 2023
According to ABB’s own literature on the Ability platform (abb.com, accessed January 2025), the IoT ecosystem requires a dedicated control backbone. It is not an afterthought. Industry-standard practice now requires planning sensor and control placement at the same stage as fixture layout. I did not do that. It cost us a week of rework.
The ‘ABB IoT Platform Connect’ Mistake (September 2022)
I approved a quote for “ABB IoT platform connect ready” fixtures. The vendor assured me everything was compatible. I didn’t verify which generation of the gateway they were referencing. ABB’s product line has evolved. Gen 1 gateways were discontinued in late 2021 (Source: ABB Product Discontinuation Notice, 2021). The fixtures I ordered were listed as compatible with Gen 1.
The mistake: I checked the specs. I did not check the revision history. The fixtures were physically installed, but they simply couldn’t achieve the connectivity profile I’d promised to the warehouse manager.
A re-quote came back: $3,200 for gateway upgrades plus a software re-license. That was my first expensive lesson in treating IoT connectivity as a static feature set.
(Side note: I keep a list of discontinued product SKUs now. It’s not fun, but it prevents this exact scenario.)
The ‘Smart Sensor’ Reality Check (Q1 2023)
The specifications mentioned “ABB smart sensor” capabilities. I imagined: occupancy sensing, daylight harvesting, energy monitoring. All true—but the sensor placement matters enormously.
The OEM’s installation manual specified a minimum 2-meter spacing between occupancy sensors on a 20-foot ceiling to avoid false triggers. The layout I approved had sensors spaced at 1.8 meters. Result: lights that would dim unexpectedly when a forklift passed through the gap between two zones. Not a system failure, but deeply annoying to the warehouse team.
According to industry-standard print resolution requirements (300 DPI for technical drawings—a standard I was intimately familiar with from my old job), the sensor placement diagram I used was correct in pixel terms but wrong in the R3 revision. I’d reviewed an outdated version. A simple version check would have saved the headache.
After we fixed the spacing (two weekends of recabling, no fun), the system worked beautifully. The ABB smart sensor data dashboard showed an 18% reduction in lighting energy usage in the first month. But the fix cost credibility.
The ‘One-Size-Fits-All Retrofit’ Lie
Without a proper site assessment, a retrofit cannot deliver guaranteed results. This is the point I wish I’d understood six months earlier.
I had a quote that promised 60% energy savings. I’d believed the number. The actual first-month result was 18%. The shortfall wasn’t the technology; it was my assumption that the “60%” savings assumed a baseline I hadn’t verified. The existing T5 fixtures were operating at lower-than-rated wattage due to a faulty ballast. The baseline was wrong. The savings projection was technically accurate for a full-spec baseline, but I’d never asked “what baseline are you using?”
Lesson: You cannot guarantee energy savings without a site assessment. I learned this after ignoring it and making myself look foolish to my stakeholders.
What I’d Do Differently (The Checklist I Now Use)
After the third expensive mistake, I built a pre-retrofit checklist for my team. It’s saved us from at least four similar errors in the past 18 months, by my count. Here’s the core:
- Verify the IoT platform generation – Is the gateway in production? Has it been discontinued? What’s the EOL policy? (Check with ABB directly, not just the distributor.)
- Demand the sensor placement diagram – Do not approve a retrofit without a dimensioned drawing showing sensor coordinates. Verify against the OEM installation manual.
- Ask for the baseline energy reading – Not the vendor’s projection. Your own data, from the last 12 months. Only then can you evaluate the savings claim.
- Run the control wiring before the fixture wiring – The network backbone needs its own path. Do not try to piggyback on existing conduit unless it’s specifically rated for Low Voltage cable.
- Budget 15% for rework – In a 50-unit retrofit, something will not fit. The ceiling grid will be off by 2 inches. A sensor will be in the wrong bay. Book the budget.
In my first year (2020), I made the classic “spec sheet is enough” mistake. After 5 years of managing procurement (the last two of which have been heavily focused on lighting controls), I’ve come to believe that a successful IoT retrofit is 30% hardware, 40% planning, and 30% debugging. (I’ve also come to believe that anyone who skips step #2 is personally funding the debug phase.)
As of January 2025, our distribution center retrofit is functional. The ABB Ability dashboard shows 41% energy savings against the corrected baseline—not the 60% from the brochure, but a real number that we can defend. The smart sensor data has helped us identify two under-utilized zones for repurposing. The warehouse manager says it’s “fine,” which in his vocabulary means “it works, and I don’t have to think about it.”
That’s the real win. Not the perfect number on the brochure. The number you can actually deliver.
(Oh, and I should add: I still keep a copy of that $3,200 redo invoice on my desk. It makes a good paperweight, and a better reminder.)