Motion Sensor Specification Guide: Four Spec Mistakes That Wasted $34,000

· Clara Whitmore

Nine years ago, if you handed me a motion sensor data sheet, I would have matched the coverage number to the room size and moved on. I did that a lot. In 2017, in 2019, and, embarrassingly, as recently as 2022.

I'm a lighting controls specification support person. For the past nine years, I've helped distributors, contractors, and OEMs order the right switch, sensor, dimmer, and relay for commercial projects. I've also personally made—and documented—21 meaningful specification mistakes. Together they cost my employer roughly $34,000 in replacement product, expedite fees, callbacks, and one very awkward contractor lunch.

I'm not sharing this because my mistakes are unique. I'm sharing them because most of the errors weren't product failures. They were specification failures that looked like product failures.

The Problem: “Defective” Sensors Usually Aren’t Defective

In September 2022, an electrical contractor sent us back twenty-six ceiling sensors. The label on the box said defective. The contractor was furious; their customer was furious; the sensors were sitting in receiving like evidence.

We tested every single one. All twenty-six worked perfectly.

The actual problem? The spec was wrong before the sensors were ordered. The project was a call center with high cubicle partitions. The sensor was a PIR ceiling unit chosen because the catalog said it covered 800 square feet. It did cover 800 square feet—in an open room with nothing blocking its line of sight. In real life, at 10 feet mounting height, the partitions blocked the detection zones where people were sitting. Lights switched off while the room was full.

The contractor wasn't wrong to be angry. They were wrong about the cause.

That incident changed how I think about sensor specification. Twenty-six boxes of working products, plus a full week of callback labor, all because someone treated a coverage circle like a room layout.

The Deep Reasons Motion Sensor Specs Fail

You could blame that call center project on “wrong sensor technology.” But the deeper issues were the ones I keep seeing on orders, submittals, and cut sheets. They aren't sexy. They cost a lot.

1. Coverage circles aren’t rooms

Most buyers ask: “How many square feet does it cover?” That's the wrong question. The right question is: “What can it actually see from the mounting height in this space?”

PIR sensors detect changes in infrared energy. They need a clear view of the occupant. If a sensor is mounted above a rack, a partition, or even a row of tall shelving, the coverage map on the data sheet stops meaning much.

I once had a project in a warehouse with 14-foot aisles. The spec called for ceiling-mounted PIR sensors, because the coverage numbers looked generous. In practice, the sensors could see the aisle directly below them, but not the workstations tucked behind racking. The lights turned off while people were still working three bays over. It took us two site visits, a new sensor layout, and an embarrassed phone call to explain why “open area coverage” didn't mean “racked warehouse coverage.”

If the spec doesn't account for furniture, partitions, racking, or mounting height, the square footage number is mostly decoration.

2. The control mode doesn’t match the space or the code

Here's a distinction that causes more return labels than anything else: occupancy mode versus vacancy mode.

  • Occupancy mode turns lights on automatically when someone enters, then off when the space is empty.
  • Vacancy mode requires someone to turn the lights on manually. The sensor only turns them off.

Owners often want automatic-on because it feels modern. But in many commercial spaces, the applicable energy code determines whether automatic-on is even allowed. The version of ASHRAE 90.1 or California Title 24 that applies to the building is the real spec. Not the product catalog.

We had a project fail a final inspection because the sensors were configured for auto-on occupancy in a space where the code expected manual-on vacancy behavior. The sensor didn't care. The building department did. We swapped the configuration, paid for a re-inspection, and absorbed a three-day delay that should never have happened.

Another version of this mistake: leaving the factory default time delay untouched. A sensor that defaults to 15 or 30 minutes can be completely wrong for a space where the owner wants a two-minute timeout. Nobody writes the delay on the submittal. It causes complaints for months.

3. The relay is rated for resistive load, not for LED reality

This is the one that still makes me wince.

Motion sensors contain relays. Those relays carry a printed load rating—often something like 500W at 120V. That rating was historically based on incandescent or resistive loads. LED drivers are not resistive loads.

LED drivers draw inrush current when they power up. That inrush can be many times the steady-state current. A row of LED fixtures, or a group of track heads from a track lighting catalog, can look fine on paper—the total steady-state wattage sits comfortably below the relay rating—and still destroy the relay over time.

I saw a showroom installation where eight LED track heads were switched by one occupancy sensor. Steady-state load was nothing. But every cold start sent an inrush spike through the relay. Within three months, the sensor contacts were welded closed. Lights stayed on all night. The owner wanted to blame the “cheap sensor.” The sensor wasn't the problem. The spec was ignoring the driver.

Since then, I check the LED driver spec sheet before I trust any relay rating. If there's any doubt, I recommend a properly rated power pack or lighting contactor. That's not an upsell. It's cheaper than a callback.

4. Nobody asked what code applies

Lighting controls are installed in buildings, and buildings are regulated. But I've approved plenty of orders where the sensor model was selected without ever asking which energy code the project had to meet.

The question isn't whether your city has adopted ASHRAE 90.1 or Title 24 or IECC. The question is which version, and whether the sensor's behavior satisfies it. Some spaces require occupancy sensors. Some require vacancy sensors. Some require sensors with specific timeout limits. Some spaces don't require a sensor at all—they require a schedule-based controller.

Every manufacturer publishes compliance guidance. But the person placing the order still has to read it. A Legrand light switch with a built-in sensor can be the right product and still produce a failed inspection if the specified control behavior doesn't match the jurisdiction. In my experience, the product is rarely the offender. The missing question is the offender.

What a “Small” Spec Mistake Actually Costs

It's tempting to think the cost of a wrong spec is just the price of the product. It isn't.

Let me give you a real example. On the call center project:

  • The sensors themselves were about $3,100.
  • Return freight, testing, and restocking: $900.
  • Electrician troubleshooting time before we got involved: roughly $2,400.
  • Second visit after the re-spec: $1,800.
  • Contractor trust: priceless, and also gone.

Total was around $8,200 for a problem that a five-minute conversation about partitions would have prevented.

Then there were the soft costs. The contractor told another contractor we sold unreliable equipment. The end user almost switched brands. The manufacturer didn't cause the failure, but the manufacturer's name absorbed the blame. That's the part that never shows up on a profit-and-loss statement, and it hurts the most.

I've calculated that the 12-point checklist I use now has prevented roughly $18,000 in potential rework over the last 18 months. The checklist isn't clever. It's just cheaper than the mistakes.

What I Do Now: A Short Pre-Order Checklist

The detailed version lives on our team's server. The short version fits on a sticky note, and honestly, it should be printed on every sensor order form:

  1. Walk the space, don't just measure it. Mounting height, partitions, racking, and ceiling obstructions change coverage. A coverage circle from a data sheet is not a room layout.
  2. Decide the behavior first. Occupancy or vacancy? Auto-on or manual-on? What timeout delay actually serves the user?
  3. Check the load, not just the wattage. LED drivers, track heads, and electronic ballasts have inrush. If the sensor relay isn't rated for the real load, add an appropriate power pack or contactor.
  4. Ask which energy code applies. Then confirm the sensor configuration matches it. This is not the manufacturer's job; it's the specifier's job.
  5. Write it down. A sensor that isn't documented, commissioned, and labeled is a sensor that will be reconfigured by someone else later.

If you're using a Legrand lighting system, or any lighting controls manufacturer's system, the same rules apply. The brand won't save you from an incorrect assumption. The best product installed under the wrong specification still fails.

My experience is based on roughly 300 commercial lighting control orders and submittals, mostly in office, retail, and light industrial spaces. If you're working in a specialty environment—hazardous locations, perhaps, or unusual ceiling heights—your experience might differ. That's exactly the point. The spec is not the product. The spec is the thought process around the product.

I still make mistakes. I make far fewer of them now, and the ones I make get caught before they become a box of “defective” sensors in receiving. Five minutes of verification beats five days of correction. It took me $34,000 to learn that. Hopefully it costs you a lot less.