You are not permitted to download, save or email this image. Visit image gallery to purchase the image.
The greenshell mussel industry, largely based out of the Marlborough and Tasman regions, has to rely on collecting its juvenile mussels, called spat, from the wild when they wash up on kelp on Ninety Mile Beach.
The industry spends more than $8 million a year collecting kelp by hand off the beach before it is placed on ropes for the mussels to grow in the Marlborough Sounds.
However, 95% of the spat captured are lost, limiting the growth of the industry into overseas markets.
Techion is based on the Invermay campus and specialises in imaging technology for disease management in the agriculture sector.
The project, which has just finished its first phase, has redesigned Techion’s imaging platform to be used on mussels.
Techion’s device would be able to tell how many mussels there were on one of the ropes and tell whether they were green or blue mussels, the company’s managing director, Greg Mirams, said.
The blue mussels were considered a pest in the industry and had no value.
Using Techion’s platform, a mussel farmer would also be able to assess the health of a mussel and make decisions, such as reseeding or shifting them for further growth.
The industry was currently working blind and there was a lot of "guess work", Mr Mirams said.
"It is just a feel and guess thing at the moment. They have to predict what their yield is going to be until they can see it.
"What this does is makes an invisible problem visible and they can make much better management decisions because of it," he said.
Designing the new platform was a "comfortable pivot" for Techion because many of the problems the mussel industry was facing, the farming sector was also facing as well.
"They want reliable information and they want it from the field, so instead of sending something to a lab, it can be done on the beach or on the barge and that is a massive breakthrough for the industry," Mr Mirams said.
The project was now moving into stage two, which would involve developing artificial intelligence to analyse the data in real time.
By late next year the company hoped to have a commercial product ready for the industry, Mr Mirams said.