Partner Portal
AI Vision  for Nursery Operations  at Driscoll’s

AI Vision for Nursery Operations at Driscoll’s

Client

Driscoll’s is one of the world’s leading producers of fresh berries, operating across 20+ countries and partnering with thousands of growers globally.
The company focuses on delivering high-quality plants and maintaining consistent production standards across its operations.

About the project

Driscoll’s needed accurate, scalable visibility into nursery inventory, as manual tray counting and dropout tracking led to inconsistencies and operational inefficiencies at the very start of the revenue chain.
To address this, we implemented the Tray & Dropout Counting solution, automating tray counting and plant loss detection using computer vision and boom-mounted cameras.

  • Client: Driscoll’s Australia
  • Solution: Tray & Dropout Counting
  • Crop: Strawberries
  • Data Collection: GoPro cameras mounted on irrigation boom

Business challenge

Driscoll’s nursery operations represent the starting point of the entire revenue chain, where inaccuracies in plant inventory directly impact downstream yield, supply planning, and overall business performance.
However, the lack of precise, scalable tracking mechanisms made it difficult to maintain consistent visibility into tray counts and plant losses, limiting the ability to make data-driven operational decisions.

Key Problems

Key Problems
  • Manual inventory tracking

    Tray counting and dropout identification relied on manual processes, leading to inconsistencies and human error

  • Manual inventory tracking

    Tray counting and dropout identification relied on manual processes, leading to inconsistencies and human error

  • Manual inventory tracking

    Tray counting and dropout identification relied on manual processes, leading to inconsistencies and human error

  • Manual inventory tracking

    Tray counting and dropout identification relied on manual processes, leading to inconsistencies and human error

Solution

AI-Powered Tray & Dropout Counting


We implemented our vision system for Tray & DropOut counting, utilizing cameras installed on the irrigation boom. This setup allows for continuous data collection across the entire table width. To ensure comprehensive coverage, the cameras operate with a 20-30% overlap between adjacent fields of view.

A key component of the solution is the use of QR codes to uniquely identify plant batches, rows, and benches. These codes link the visual data directly to critical information, including Batch ID, Plant Type, and Plant Variety.

Our technology processes the video data to detect individual trays and highlight dropouts, providing an automated and highly accurate census of the nursery inventory.

solution

Results

The deployment provided Driscoll’s AU with immediate, verifiable data on plant inventory and losses, moving beyond estimates and manual counts:

MetricResult
Mean Percentage Error (Tray Count)1.5%
Mean Percentage Error (Dropout Count)1%
drone

This unprecedented level of precision allows nursery managers to utilize a digital Dashboard and Tray Map visualization.
The Tray Map specifically colors areas based on dropout numbers, enabling quick identification of trays with 1–5 dropouts (yellow) or more than 5 dropouts (red), facilitating timely intervention.

Future Plans

The solution has become an important part of Driscoll’s operations and helped build a strong ongoing partnership.
With accurate inventory data at the nursery level, Driscoll’s can better plan supply and consistently provide high-quality plants to their growers. This strengthens the foundation of their entire revenue chain.
In simple terms, the system replaces manual counting and rough estimates with clear, precise data, giving teams confidence in their decisions and visibility into what’s happening in real time.

Reach out to see how we can grow your business smarter.

AI Vision for Nursery Operations at Driscoll’s