Source.ag updates tomato harvest forecast model with 33% accuracy improvement at three-week horizon

Dutch startup Source.ag cut tomato harvest forecast errors by 33% at the three-week mark with its updated AI model. It also halved severe prediction misses and reduced manual data entry for growers.

Categorized in: AI News Management
Published on: Mar 25, 2026
Source.ag updates tomato harvest forecast model with 33% accuracy improvement at three-week horizon

Dutch Greenhouse Software Improves Tomato Harvest Forecasts by 33%

Source.ag, a Dutch platform that uses AI to forecast harvests for greenhouse growers, has released an updated model for tomato production with measurably higher accuracy. The new model reduces manual data entry for growers while improving prediction performance across the board.

At the three-week forecast horizon, mean accuracy improved 33% compared with the previous version. The company also cut by half the number of severe forecast misses - cases where predictions missed targets by a large margin.

How the Model Works

The model combines horticultural science with machine learning that adapts to each greenhouse's specific conditions. It accounts for climate variability, equipment differences, and operational factors that affect daily yields.

The key change: the model now learns from every cultivation running through Source.ag's platform. As more growers use the system, the model's predictions improve automatically - without requiring manual updates from the company.

Less Work for Growers

Source.ag automated several data processes that previously required regular manual input from growers. The reduction in administrative work removes what the company describes as "real friction from the grower's week."

Sebastiaan Vermeulen, data scientist at Source.ag, said the changes make it "easier than ever before to hit harvest commitments, realize a better price and reduce waste."

Rollout Timeline

A small group of greenhouses are already running the new model. Full deployment to all tomato customers is expected within months.

Rien Kamman, CEO and co-founder of Source.ag, said the model represents the beginning of a broader shift toward AI that improves with scale rather than manual effort alone.

For managers overseeing agricultural operations, the practical takeaway is clear: better forecasts with less administrative overhead. Those responsible for supply chain planning or harvest logistics may want to track how these accuracy gains translate to reduced spoilage and improved delivery reliability.


Get Daily AI News

Your membership also unlocks:

700+ AI Courses
700+ Certifications
Personalized AI Learning Plan
6500+ AI Tools (no Ads)
Daily AI News by job industry (no Ads)