AI pineapple picker moves closer to reality
What started as a grower's idea to improve pineapple harvesting is becoming a working reality, with growers getting a firsthand look at the latest prototype during the 2026 Pineapple Field Days held recently in Hervey Bay.
Wide Bay grower Jeff Atkinson has never been afraid to look beyond traditional farming practices. Drawing on experience across sugarcane, soybeans, cotton and pineapples, Jeff has spent the past several years adapting machinery and exploring automation to solve one of the industry's biggest challenges – harvesting fruit at exactly the right time.
Working alongside Brisbane-based industrial automation and robotics specialists AKA Electrical & Automation, Jeff is developing an AI-powered vision system capable of detecting, and ultimately, selecting pineapples based on their maturity.
The prototype was demonstrated to growers during the Pineapple Field Days, providing a glimpse of how artificial intelligence and automation could one day support pineapple production.
Ethan Harvey-Davidson from AKA Electrical & Automation explained the current system uses onboard cameras and AI vision technology to identify whether a pineapple is present before activating a mechanical paddle that harvests the fruit and transfers it onto a conveyor.
"We have two cameras because the machine operates across two lanes," Ethan said.
"As each camera detects a pineapple, it activates a paddle that snaps the fruit from the plant, transfers it onto the conveyor and moves it through the machine."
While simply detecting fruit is an important milestone, the next phase of development is where the technology could deliver significant value for growers.
The team is now working towards teaching the AI system to recognise fruit colour, allowing growers to adjust harvest settings from a touchscreen display in the cab.
"Jeff wants to be able to use a sliding scale so the system can identify the colour of the pineapple and only harvest fruit at the desired stage of maturity," Ethan said.
For pineapples, timing matters. Unlike many fruits, pineapples do not continue to ripen after harvest, making accurate harvest timing critical to delivering the sweet eating experience consumers expect.
The technology also has the potential to help address labour challenges while improving harvesting efficiency.
Although the current prototype still requires an operator to drive the machine, Ethan said the long-term vision is to integrate the harvesting system with autonomous vehicle technology.
"Once we have colour detection working reliably, we'll have a very solid product," he said.
"From there, the goal is to attach the system to an autonomous vehicle so it can eventually harvest paddocks with minimal operator input."
Beyond harvesting, the AI platform could also unlock valuable on-farm data.
The vision system can already count harvested fruit, with future development expected to include quality assessment, defect detection and even in-field grading before fruit reaches the packing shed.
"It's the same type of industrial automation technology we use in factories," Ethan explained.
"It has the capability to do quality assurance, identify defects, measure size and provide growers with much more information than simply whether there's a pineapple there."
For Jeff, the project has always been about finding practical solutions that improve productivity while helping the pineapple industry continue to innovate.
As development continues over the coming season, the prototype is another example of how growers are leading agricultural innovation - combining practical farming knowledge with emerging technology to solve real-world challenges and build a more productive, resilient future for Queensland horticulture.
