Empowering Luminaires

Although I feel like I'm repeating myself, it seems the concept of distributed lighting control is so new, that we need to talk about it over and over again. This whole concept builds on the strengths of Bluetooth mesh: addressable information and concurrent multicast (many-to-many) communications. And the whole idea can be implemented thanks to the Moore's Law: it is possible today to move the control logic and algorithms to the edge nodes.

Making the central "control box" obsolete.

In lighting control systems this approach is disruptive. It renders the whole category of products - legacy lighting controls - obsolete.

A legacy lighting system consists of sensors (they may be standalone - room level, or integrated into fixtures - luminaire level). The second building block is the controller - a box (a computer) that collects information from the sensors, processes and sends control commands to drivers that in turn change the output, dimming the lights down or up.

Lighting control is a software function. There is nothing special about the controller - it is typically an embedded computer that runs lighting control software. And this software can be relocated to the luminaires (typically can be integrated into drivers - for room level sensing or into sensor-controllers - for luminaire-level sensing).

The result is a system that consists only of sensors and luminaires.The controller is gone. And so is the vendor that provides that whole thing, has products, a roadmap, distribution, training, and - of course - adds cost to the system.

This transformation may be illustrated as an electric car that has an electric motor in each wheel: the central engine, the gearbox, the drive shaft, the differential, are all gone. The wheels suddenly are no longer passive elements of the car, as each has its own integrated motor, reacting to input from the driver as well as other sensors. So are the lights in a distributed system - they have become smart lights, independently capable of interpreting sensory information and adjusting the behavior. And as in the car example, this architecture radically simplifies the whole setup: no central engine, no central gearbox, no drive shaft. And no single point of failure: if one wheel breaks, the others keep on turning.