How artificial intelligence is being used to weed out bad strawberries

THE farms using this new technology were not implicated in the needle contamination saga, so is this the answer we need?

AN ADVANCED piece of technology used to weed out bad strawberries and prevent human error could be the answer to saving the industry.

Griffith University on the Gold Coast has developed an artificially intelligent machine that makes sure all berries passing through a conveyor are of the same high standard.

Professor Yongsheng Gao said while their technology was not designed to detect metal - because such devices are already in use - it picks up all kinds of foreign objects, such as debris, insects or bandages that have accidentally slipped in during the packing process.

None of the three farms the university is working with - two in Queensland and one in South Australia - have been implicated in the current needle contamination saga.

A nationwide scare involving the piercing of various fruits with sewing needles has prompted a series of supermarket recalls, and some stores in New Zealand have temporarily banned the sale of Australian strawberries.

Police have received reports of more than 100 alleged cases of pins and needles being found in strawberries, bananas and mangoes, since the first incident was recorded in Queensland on September 9.

On Friday, a contaminated strawberry with a needle inside was found in the Northern Territory, the first case of its kind for the state.

Where it all began: The original needle found in a strawberry in Brisbane.

At least two minors have so far been questioned by police but authorities have struggled to find the original culprit as a spate of copycat episodes threaten to bring the industry to its knees.

Sewing needles were taken off the shelves at major Australian supermarket chain Woolworths on Thursday in response to sharp objects being found inside strawberries and other fruits across five states.

Professor Gao said the technology would ensure high quality and consistent strawberries.

It detects unripened, undersized and bruised fruit.

“There is definitely a need from the industry,” he said.

“Inconsistency of different workers results in inconsistent fruit quality and this machine will help maintain a high standard of consistency.”

Professor Gao said farms would be able to increase prices, prevent loss and have handlers working less tediously because they would instead be operating a machine to check berries.

The strawberry robot in use at a farm. Picture: Griffith University

He said currently there was no way of figuring out which worker was responsible for bad batches rejected by supermarkets.

“Our system can identify the linkage so that person can be retrained instead of all 300 workers being blamed for the error of one,” he said.

“It also encourages good behaviour because it’s monitored.

“It’s a deterrent for people (tampering with fruit). This system is progressively evolving but if the strawberry is tampered with or abnormal, it will detect it.”

Professor Gao said the biggest problem to detect was fresh bruising because it was often not spotted until it got to supermarkets where it was rejected.

The project has been given $5 million in funding from the Australian Government.