For over 40 years, cannabis breeders and cultivators have sought ways to identify a seed's traits before it even sprouts. Now, thanks to advancements in spectral imaging, this vision is becoming a reality.
INNEXO, in collaboration with Innoveins Seed Solutions and SeQso, has successfully completed and validated a proof-of-concept study, elucidating the potential for AI-driven seed sorting to cannabis for the first time ever.
This cutting edge technology, already validated in other high-value crops such as watermelons, sugar beets and tomatoes, has now been adapted to identify and correlate spectral signatures of cannabis seeds with critical traits, including sex, off-types, germination rate and flowering characteristics.
"When we at INNEXO first heard about the validated success of this technology in other crops, we sat out to prove that this could not be translated to cannabis. Because we knew that if it could, the implications for the cannabis industry would be unimaginable. Never was I more excited to be disappointed." - Dominique van Gruisen, CEO of INNEXO
A new standard for quality
Through advanced spectral imaging and AI-powered machine learning, INNEXO and its
partners have developed a non-destructive, high-throughput method to analyze and sort cannabis seeds based on genetic and phenotypic traits.
In translating this advanced technology to cannabis seeds, a trained and validated model was used from other crops. Seeds were screened by a spectral camera array on a large data collector. 600 Seeds of 6 different cultivars were digitally labeled, and over 5.000 datapoints were collected per individual seed. The subsequent grow-out was performed under controlled conditions and a variety of traits were meticulously annotated. After statistical modelling of the annotated traits with the spectral dataset, the resulting traits that have sorting viability were identified.
The results of this proof-of-concept show validity for germination viability, sex sorting, off-type elimination and phenotypic trait selection.
"At INNEXO we create proprietary sorting algorithms for our clients such as breeders and seed companies. Because every cultivar carries a unique spectral signature, and the techniques we use are non-destructive, proprietary sorting algorithms can be developed for every cultivar in our clients portfolio. Together with our industry leading partners, we can deliver a tailored solution that is accurate, scalable and economically viable. What is even more important: as an industry we don't need to wait ten years to make this technology the new benchmark. We are ready to go." said Dominique van Gruisen, CEO of INNEXO.
For more information:
Innexo BV
[email protected]
www.innexo.nl