Smarter Lights, Smaller Bills: AI Enters the Greenhouse

Sollum Technologies launches SF-INFINITE, an adaptive LED system that uses AI to adjust greenhouse lighting in real time

6 Mar 2026

A Canadian horticultural lighting company has launched an adaptive LED fixture that uses artificial intelligence to adjust growing conditions in real time, aiming to reduce the operational burden on greenhouse operators who have long relied on fixed infrastructure.

Sollum Technologies introduced the SF-INFINITE on February 25 in Leamington, Ontario, at the Greenhouse Grower Expo. The Montreal-based firm's system connects individual light fixtures to a cloud platform that continuously modifies spectrum, intensity, and timing across crop zones. Up to four independently controlled channels operate per fixture, allowing growers to run separate programmes within the same facility without hardware changes.

The system is built to respond to external pressures as well as internal ones. When energy prices shift, the platform dims output automatically to protect margins, in what the company describes as tariff-aware dimming. When a new crop enters production, lighting parameters are updated through software. An edge-computing function embedded in each fixture maintains operations if cloud connectivity is lost. Firmware updates are delivered remotely over time.

The launch comes as the global greenhouse and vertical farming sector, valued at roughly $7.5bn in 2026, faces mounting pressure to improve energy efficiency without expanding physical footprint. Energy remains the largest variable cost in controlled environment agriculture, and operators are increasingly expected to do more with existing infrastructure.

Sollum's co-founder framed the product's underlying logic in functional terms, describing the fixture as "a computer on the greenhouse ceiling" that learns from energy data, weather signals, and crop performance over time.

The SF-INFINITE was demonstrated at the Leamington expo across tomato, cucumber, pepper, and propagation production, crops that collectively represent a significant share of Ontario's greenhouse output.

Whether the platform delivers on its promise of long-term adaptability will depend on how well it performs across the varied conditions of commercial-scale production. For an industry accustomed to hardware replacement on fixed cycles, the shift to software-defined infrastructure carries both appeal and uncertainty. Adoption rates and grower outcomes over the near term will test the proposition.