Most industrial LED retrofit projects stop at swapping the fixtures. New lights go up, energy drops, everyone declares victory. Six months later the lights are running at full output in an empty warehouse at 2 a.m. because nobody thought about controls.
That missed piece — the control layer — is where another 40 to 70 percent of lighting energy goes. Fixtures get you the first cut. Controls get you the rest. And in a 200,000-square-foot facility burning through $80,000 a year in lighting electricity, the gap between “we changed the bulbs” and “we actually automated the system” is tens of thousands of dollars annually.
This guide walks through the control protocols, sensor technologies, and integration approaches that turn a basic LED retrofit into an intelligent lighting system — without the consultant-speak.
The Three Control Protocols That Matter in Industrial Settings
Industrial lighting controls have consolidated around three standards. If you’re specifying a system today, these are the ones you’ll encounter.
DALI-2: The Standard That Finally Works
DALI (Digital Addressable Lighting Interface) has been around since the late 1990s, but the original standard was a mess — partial interoperability, vague testing requirements, and every manufacturer’s idea of “DALI-compatible” meant something different.
DALI-2, standardized under IEC 62386, fixed that. Every DALI-2 certified device passes a mandatory interoperability test. A DALI-2 driver from Manufacturer A talks to a DALI-2 sensor from Manufacturer B without custom integration work. That was the promise of the original DALI. DALI-2 actually delivers it.
Key capabilities:
- Individual fixture addressing: Each luminaire gets a unique address on the bus. You can dim, group, and scene-control individual fixtures or zones of fixtures independently.
- Two-way communication: The controller tells a fixture to dim to 50 percent. The fixture reports back its actual status, energy consumption, and any faults. You know when a fixture is failing before someone files a maintenance ticket.
- Sensor integration on the same bus: DALI-2 supports occupancy sensors and daylight sensors as native bus devices. No separate sensor wiring loop.
- 64 devices per bus: Each DALI line supports up to 64 devices (drivers, sensors, switches) with a maximum bus length of 300 meters using standard 1.5mm² wiring.
The tradeoff: DALI-2 requires a dedicated two-wire control bus, separate from mains power. In a new-build facility, that’s trivial — you run control wiring alongside power. In a retrofit, pulling new control wire through existing conduit or across finished ceilings adds labor cost. Whether that cost justifies the capability depends on the facility’s complexity and how much granular control you actually need.
0-10V: Simple, Ubiquitous, Limited
0-10V analog dimming has been the default in commercial and industrial lighting for decades. It works by varying a DC voltage between 0 and 10 volts on a dedicated pair of control wires: 10V = full brightness, 1V = minimum dimming, 0V = off (if the driver supports it).
What 0-10V does well:
- Nearly every LED driver supports it
- Installation is well-understood by every electrical contractor
- Sensors and wall switches are cheap and widely available
- No configuration or commissioning required — it’s analog, it just works
What 0-10V does poorly:
- No two-way communication. You dim the fixtures, but you don’t know if they actually dimmed.
- No individual addressing. All fixtures on a 0-10V circuit move together.
- No fault reporting. A dead driver is invisible to the control system.
- Voltage drop over long runs shifts the dimming curve. A fixture at the end of a 200-foot control wire run may be at a different brightness than the one at the start.
For simple applications — a warehouse aisle where all fixtures in a zone dim together based on one occupancy sensor — 0-10V is perfectly adequate. For a multi-zone manufacturing floor where different production cells need independent lighting profiles, DALI-2 or wireless is the better choice.
Wireless Mesh: Thread, Zigbee, and Bluetooth Mesh
Wireless lighting control eliminates the control wiring entirely. Each fixture includes a wireless radio that forms a mesh network with neighboring fixtures. Commands hop from node to node across the network, so a sensor at one end of the facility can control fixtures at the other end without direct line of sight to a gateway.
Three protocols dominate:
Thread (based on IEEE 802.15.4, the same radio as Zigbee but with IP-native networking): The newest entry, backed by the Thread Group (Google, Apple, Amazon, Samsung). Thread builds an IPv6 mesh network where every fixture is addressable over standard IP protocols. For facilities that already have an IP-based building management network, Thread eliminates the protocol translation layer that Zigbee requires. Adoption is growing but the ecosystem of industrial-grade Thread lighting products is still smaller than Zigbee’s.
Zigbee (Zigbee 3.0, also 802.15.4): The incumbent. Thousands of compatible products from dozens of manufacturers. Zigbee mesh networks self-heal — if one fixture loses connection, traffic routes around it. Range between nodes is typically 30 to 50 meters indoors, which means even a sparsely populated warehouse needs fixtures spaced close enough to maintain the mesh. In steel-framed buildings with lots of metal racking, radio performance can degrade unpredictably. A site survey before committing to a wireless-only design catches most of these issues.
Maillage Bluetooth (BLE 5.0 and later): The lowest barrier to entry because commissioning can be done from a smartphone app. No gateway required for basic operation — a maintenance tech walks the floor with a phone and provisions each fixture directly. For facilities with 50 to 200 fixtures, this is often the simplest path. For facilities with 2,000 fixtures spread across multiple buildings, the smartphone-as-commissioning-tool approach becomes tedious, and a gateway-backed architecture makes more sense.
A practical rule of thumb: if your facility has fewer than 500 fixtures and open sight lines, wireless works well. If you have more fixtures, dense metal racking, or areas that need guaranteed sub-200ms response times (like safety-critical zones), hardwired DALI-2 is the safer bet.
Sensor Types and Where to Use Them
PIR (Passive Infrared) Occupancy Sensors
PIR sensors detect changes in infrared radiation — basically, they see warm bodies moving through their field of view. They’re cheap, reliable, and everywhere. The limitation: PIR requires line of sight. Hide behind a warehouse rack and the sensor won’t see you.
Best for: open areas, aisles, corridors, and break rooms where people are moving through rather than sitting still.
Microwave / High-Frequency Doppler Sensors
These emit a low-power microwave signal and detect the frequency shift when it bounces off a moving object. They penetrate non-metallic barriers — a microwave sensor mounted in a fixture can see movement through plastic diffusers and even thin partition walls. They’re more sensitive than PIR, which is both a feature and a problem. A vibrating conveyor belt or a fan blowing a loose cable can trigger false positives.
Best for: high-bay areas, production floors with partial obstructions, and locations where PIR’s line-of-sight limitation is a dealbreaker.
Dual-Technology Sensors (PIR + Microwave)
Both sensors must trigger for the lights to turn on. Either sensor alone keeps them on. This combination eliminates false triggers from the microwave side (the PIR won’t confirm the vibration) while avoiding the “lights go out while you’re standing still” problem of PIR-only sensors.
Best for: offices, labs, and areas where people may be stationary for extended periods.
Daylight Harvesting Sensors
These measure ambient light and adjust fixture output to maintain a consistent illuminance at the work surface. Near windows and skylights, fixtures dim during daylight hours and ramp up as natural light fades. In a facility with 20 percent glazing area, daylight harvesting typically cuts lighting energy by another 15 to 25 percent on top of occupancy-based savings.
The catch: daylight harvesting only works if the sensor is positioned correctly and calibrated after installation. A sensor mounted directly above a skylight will read sky brightness, not task-level brightness, and dim the fixtures when they should be at full output. Calibration during commissioning — measuring actual lux at the work surface and mapping it to sensor readings — is the difference between “daylight harvesting works” and “someone disabled it after the first week because the lights kept flickering.”
Integration with Building Management Systems
Industrial facilities increasingly run on centralized building management systems (BMS) or industrial IoT platforms. Lighting controls that can’t talk to the BMS become yet another silo that someone has to monitor separately.
BACnet is the common language of building automation. A DALI-2 system with a BACnet gateway can expose individual fixture status, zone energy consumption, and fault alerts to the BMS that already monitors HVAC, access control, and fire safety. The facility manager sees everything on one dashboard instead of juggling three separate systems.
MQTT is gaining traction in newer industrial IoT deployments. It’s a lightweight publish-subscribe protocol designed for sensor data. A Thread-based wireless lighting network can publish occupancy data directly to an MQTT broker. That occupancy data is useful beyond lighting — the HVAC system can reduce airflow to unoccupied zones, and the security system can log unexpected movement in areas that should be empty at 3 a.m.
Energy Savings: What Controls Actually Deliver
Independent studies from the DesignLights Consortium and Lawrence Berkeley National Laboratory provide real-world savings data for industrial lighting controls:
| Control Strategy | Typical Energy Savings | Best Application |
|---|---|---|
| Occupancy-based on/off | 20-35% | Warehouses, storage areas, corridors |
| Occupancy-based dimming (dim to 20%) | 35-55% | Production floors, logistics centers |
| Daylight harvesting | 15-25% | Perimeter zones, skylit areas |
| Task tuning (reduce max output to design target) | 10-20% | Over-lit retrofit spaces |
| Combined occupancy + daylight + task tuning | 50-70% | Full-system deployment |
Task tuning deserves attention because it’s the lowest-hanging fruit in most retrofits. When an old 400W HID fixture gets replaced with a 150W LED fixture, the LED often puts out more usable light than the HID did on its best day — because LED optics put light where it’s needed instead of spraying it everywhere. Dialing the maximum output down to the actual design illuminance target (rather than running at full blast) saves energy with zero impact on lighting quality and takes 30 seconds per fixture during commissioning.
How to Specify a Control System Without Over-Engineering It
The most common mistake in industrial lighting control projects is specifying more capability than the facility will ever use. Individual fixture addressing with scene-based control sounds impressive in a proposal, but if the facility runs the same lighting schedule every day and nobody ever adjusts individual zones, you’ve paid for complexity that delivers zero additional savings.
Start here:
- Define the zones: Walk the facility and mark which areas need independent control. A warehouse with 10 aisles that all operate on the same shift schedule probably needs one zone. A manufacturing floor with three shifts, each occupying different production cells, probably needs three to six zones.
- Pick the dimming strategy: On/off is simpler than dimming. Dimming saves more energy but adds cost. In aisles and storage areas, on/off occupancy control is usually sufficient. In occupied production areas where sudden darkness is a safety hazard, dim-to-20-percent is the safer choice.
- Choose wired vs. wireless: New construction → run DALI-2 control wiring, it’s a marginal cost add. Retrofit → evaluate wireless mesh. If the facility has dense metal racking, get a site survey and a pilot installation in one zone before committing to wireless across the entire facility.
- Plan for commissioning: The best control system in the world is worthless if it was never properly commissioned. Budget for one to two days of commissioning time per 500 fixtures. That includes sensor calibration, zone programming, dimming curve verification, and occupancy sensor sensitivity adjustment.
- Document the configuration: Print a zone map, label every controller, and leave a copy in the electrical room. The next facility manager — the one who wasn’t there during installation — needs to understand how the system works without calling the original contractor.
Questions fréquemment posées
Is DALI-2 worth the extra cost over 0-10V for a retrofit?
It depends on the facility’s complexity. For a single-zone warehouse where all lights operate on the same schedule, 0-10V with a few occupancy sensors is perfectly adequate and costs less. For a multi-zone manufacturing facility where different areas run different shifts and fault monitoring matters for maintenance planning, DALI-2’s two-way communication and individual addressing justify the cost. If you’re unsure, cost out both options — the DALI-2 premium is typically 15 to 25 percent on the controls hardware, with most of the delta coming from the control wiring labor.
Can wireless and wired controls coexist in the same facility?
Yes, and this is increasingly common. Use DALI-2 in the production floor where response time and reliability are critical, and wireless BLE Mesh in office and break areas where the lower installation cost matters more. A BACnet-capable gateway that supports both protocols ties them together for centralized monitoring.
How long do occupancy sensors take to turn lights on and off?
Turn-on response should be under 500 milliseconds — the light should be at full brightness before someone notices it was ever off. Turn-off delay is configurable; 5 to 15 minutes is typical for occupied areas to avoid nuisance cycling. Shorter delays (1 to 3 minutes) work in transient spaces like corridors and restrooms.
Do I need an IT network for a wireless lighting control system?
Most wireless lighting systems operate on their own mesh network and don’t touch the corporate LAN. The gateway device that bridges the lighting mesh to the building management system does need a network connection, but lighting mesh traffic stays on the lighting network. If you want cloud-based monitoring and analytics, the gateway needs internet access. If you’re fine with on-premises monitoring only, it just needs a local IP address on the BMS VLAN.
Résultat final
Lighting controls aren’t complicated technology. They’re sensors, dimmers, and a communication protocol — nothing that didn’t exist a decade ago. What’s changed is that the standards have matured to the point where multi-vendor interoperability actually works, and the hardware costs have dropped enough that controls pay for themselves within two to four years in most industrial facilities.
If you’re planning an LED retrofit, spec the controls at the same time. Adding controls later means paying for the same labor twice — once to hang the fixtures, once to add sensors and gateways. Integrated sensor-fixtures that arrive pre-wired from the factory turn what used to be a two-trade job (electrician plus controls contractor) into a single electrical installation. That’s the difference between a project that gets approved and one that stays in the budget spreadsheet.
For help specifying a control system for your specific facility layout, including zone planning, protocol selection, and ROI projections, contact the Recolux engineering team.