Source.ag’s greenhouse deal moves AI from crop forecasts to climate settings

For Mike Arts, head grower and owner at Integral Farms in Ontario, a change in electricity prices used to mean an hour updating settings in his Priva greenhouse computer. In a testimonial published by Optimal, he describes now changing his climate targets and letting the software handle the corresponding settings. The distinction captures the work Source.ag has just bought.[2]

On 9 October, the Dutch greenhouse software company announced its acquisition of London-based Optimal. Founder Dave Hunter and his engineering team are joining Source, with Hunter becoming vice-president of autonomous control. Financial terms were not disclosed. Source already monitors crops, predicts harvests and automates irrigation; Optimal adds control of the wider greenhouse climate.[1]

The greenhouse already has a computer

Modern climate computers coordinate substantial machinery. Priva’s Connext, for example, manages light, climate, water and energy systems together. Its documentation describes a gas strategy that plans heat production and carbon-dioxide supply, using a heat buffer to help reconcile their different requirements. A grower is configuring a connected production system.[3]

Closing a greenhouse screen illustrates the complication. Priva explains that it changes light, temperature and energy use, with consequences for moisture, water uptake and irrigation. Adjusting one setting can therefore change the circumstances in which several others operate. The practical task is to keep a growing strategy coherent as conditions move.[3]

Optimal’s current interface starts with the climate the grower wants. It shows predicted conditions and planned settings for each compartment over the next 24 hours, revising the plan as weather forecasts change. The company says its software sends settings through official interfaces to Priva or Ridder computers, with no additional sensors required. Growers can return to manual control.[2]

Source describes the underlying engine as a combination of physics and machine-learning models of the greenhouse and crop. Its planning covers heating, ventilation, screens, lighting, carbon dioxide and irrigation.[1]

That division of work matters commercially. The grower supplies a cultivation objective; the software carries it through many linked adjustments. A business can delegate the repeated configuration work while retaining a person who decides whether the chosen growing conditions still suit the plants. Buying this capability is a more specific proposition than buying an autonomous farm.

A harvest forecast has customers attached

Source’s Cultivation Management product lets growers simulate choices such as temperature, stem density and pruning, then follow actual crop development against the plan. Its included Harvest Forecast connects expected production with sales commitments. Growers can explore how a change might affect harvest volume, timing or fruit weight before updating the commercial team.[4]

Those records create a second reason to connect planning and control. A forecast can tell a sales team that fewer tomatoes are likely to arrive in a particular week. Climate control concerns the physical decisions that influence the crop’s progress. Putting both in one product could make it easier to follow a proposed growing change through to the quantities the business expects to sell. That is a potential benefit of the combination, rather than a reported result from the acquisition.

An existing system, with a new route to growers

Optimal has a commercial record before this deal. HortiDaily’s September 2023 reporting described its demonstration greenhouse in Westdorpe, the Netherlands, and a wider operating system that combined predictive control with procedures for human workers. Optimal reported 13% higher yield and 27% energy savings against a benchmark Dutch greenhouse. The report also described experiments with tomato and cucumber varieties, without identifying the crop mix or comparison period behind those headline percentages.[5]

Those are company-reported results from the earlier operation, not measurements of the newly combined Source product. The wider system matters: a comparison that includes growing procedures and operational supervision cannot isolate the contribution of one control model.[5]

Source says its software is already used in more than 500 greenhouses across 25 countries. That is its existing platform footprint. The first autonomous deployments at Source customers in Europe and North America are planned for this winter.[1]

The acquisition gives Optimal a route into businesses already using Source to examine their crops and plan production. For Source, it adds another practical role: helping growers carry those plans into the settings of the equipment they already own. The harvest forecast and the climate controls would become two connected views of the same growing decision.

Report an error or suggest a correction