Agtech companies around the globe are attempting to be first to market with pre-harvest, on-tree, fruit technology. Meanwhile, award winning fruit tech startup, Hectre, have quietly solved a big part of the problem already.
Hectre, the fruit tech company originating in NZ, have developed a computer vision AI solution that can size fruitlets, accurately and within seconds, without any expensive equipment.
Matty Blomfield, CEO and Co-founder of Hectre, says the idea came about through discussions with valued Hectre client, and apple innovator, Rockit Global. “Rockit has a very strong focus on innovation.”
Aware of the time involved with carrying out traditional pre-harvest manual estimations, small sample sizes, and flow on negative impacts of that manual process, Blomfield says Rockit were looking for a better way.
“Rockit have been using Spectre - Hectre’s award winning computer vision AI fruit sizing app that works straight off an iPad or iPhone - for post harvest size and color estimations and wanted to know if there was a way to use our Spectre technology on fruitlets,” explains Blomfield. “Our team built and trained a new Spectre computer vision model to work with small crates and small fruit.”
Spectre for Fruitlets customers collect fruitlets into a small crate, take a photo on iPad or iPhone, and Spectre detects and sizes a large sample of fruitlets, delivering sizing results within seconds. No manual caliper or sizing ring calculations required, saving significant amounts of time.
Caption: Spectre for Fruitlets delivers pre-harvest crop size estimation data within seconds, all from a standard iPhone or iPad.
Because the technology is so simple and quick, rapid samples can be captured across blocks and varieties over time. Weekly samples significantly enhance representation of the crop providing time-based maturity indicators for savvy growers looking to make accurate predictions in years to come.
Hectre’s post-harvest Spectre technology is a popular solution for growers and packers seeking mobile, fast and accurate, early fruit sizing solutions. Customers include fruit giants Washington Fruit & Produce, First Fresh NZ, Borton Fruit, Kershaw and Roche, just to name a few. Spectre is now being used in 11 countries around the world. The Spectre post-harvest app detects approximately 125 pieces of fruit from the top layer of the bin, sizes it and provides a fruit size estimation graph within seconds.
Growers can gain sizing data as soon as the first bin of fruit is picked, and packers can streamline their storage, packing and sales programs, saving costs and securing high value sales orders.
“The user simplicity baked into Spectre resonates with industry, and the speed, accuracy and broad capability of Spectre is second to none, making it one of the fastest, most accurate, portable fruit sizing solutions in the world. The addition of color estimation, in any lighting condition, further adds to the popularity of Spectre,” says Blomfield.
Adds Blomfield, “Our clever simplicity approach keeps user simplicity at the forefront. If we can take a complex problem like crop size estimation and solve a key piece with a solution which is simple to use and already proven, we know we’re putting value into grower’s hands.”