The most significant shift in the lighting industry this year is not a new LED chip or a higher lumen rating - it is artificial intelligence. At the 31st Guangzhou International Lighting Exhibition, held in June at the Canton Fair Complex, more than 2,800 lighting companies from 18 countries and regions filled 21 exhibition halls covering 210,000 square meters. Across the show floor, one theme dominated: AI lighting ecosystems are moving from "passive response" toward "active thinking."
For years, smart lighting meant voice commands and app control - essentially remote controls with extra steps. The new generation displayed this year is qualitatively different. Large language models and mature IoT technology now allow lighting systems to perceive environments autonomously and make decisions. Exhibitors showed AI systems that generate complete lighting scenes from a text instruction, lamps that learn a user's schedule and adjust color temperature every 15 minutes to match circadian rhythms, and urban lighting "agents" that integrate lighting control, data collection, maintenance, and emergency response in a single intelligent platform.
One manufacturer launched the industry's first dedicated AI agent for urban lighting, breaking the traditional model of passively receiving commands and giving city lighting a "brain" that unifies illumination, data, operation and maintenance, emergency, and energy storage functions. Others demonstrated AI retail lighting that recognizes the material and color of merchandise below a fixture and automatically adjusts light parameters to render products at their best.
Nowhere is this transformation more practical than in circulation spaces. Corridors, stairwells, and aisles represent a huge share of installed luminaires - and historically, a huge share of wasted energy, since these spaces are occupied only a fraction of the time. Sensor-driven LED lighting already changed the economics here; AI takes it further by learning traffic patterns, predicting usage, coordinating groups of fixtures, and balancing safety against energy goals in real time. A modern bulkhead light for corridor applications, equipped with onboard sensing and dim-to-standby capability, exemplifies the hardware side of this shift: each bulkhead light for corridor mounting heights works as an independent smart node, yet coordinates with its neighbors so occupants always walk in a pool of light that precedes them.
The supporting infrastructure is maturing in parallel. Driver power supplies have evolved from simple power-conversion units into the "intelligent nervous system" of lighting, handling data collection, protocol communication, precise dimming, and scene linkage. Widely adopted open protocols - together with new-generation semiconductor power devices - give fixture makers the building blocks for full-scenario intelligent lighting, from standalone sensors to city-scale platforms.
Policy reinforces the technology push. National programs for smart cities and green buildings continue, while regional action plans explicitly call for AI-empowered upgrades of municipal infrastructure including lighting. Consumer demand for personalized, healthy light environments accelerates adoption in the residential segment.
Industry analysts note that competition has shifted from spec-sheet parameters toward systems capability: circadian support, scene experience, energy management, and integration competence. The global smart LED lighting market is projected to reach nearly 21.4 billion US dollars by 2030, growing at 13.2 percent annually - with industrial smart lighting, boosted by automation and AI data-center construction, the fastest-growing segment.
The conclusion from this year's exhibitions is unambiguous: for lighting companies, embracing AI is no longer optional. The light itself is becoming intelligent - and the industry that produces it must do the same.
