A team of scientists in Switzerland has developed silicon-made raspberries that can help train robots to pick and handle this delicate fruit.
Harvesting raspberries can be difficult since they can easily be damaged because they are soft. Thus, they are usually harvested by human hands.
(Photo : EPFL Official YouTube Channel)
An EPFL laboratory has created a silicone twin of a raspberry in order to train a robot to pick it with the required delicacy.
Testing Fruit-Picking Robots With Fake Raspberries
Engineers at Ecole Polytechnique Federale de Lausanne's (EPFL) Computational Robot Design & Fabrication (CREATE) lab are training robots to pick raspberries on a silicone version.
According to Interesting Engineering, the fake raspberries developed by engineers mimic the real thing. This innovation will lessen food waste in testing and avoid challenging harvesting seasons.
CREATE professor Josie Hughes shared her excitement with this invention. Hughes said in a statement, "It's an exciting dilemma for us as robotics engineers. The raspberry harvesting season is so short, and the fruit is so valuable, that wasting them simply isn't an option."
She added: "What's more, the cost and logistical challenges of testing different options out in the field are prohibitive. That's why we decided to run our tests in the lab and develop a replica raspberry for training harvesting robots."
Their study was posted in this week's journal Communications Engineering. The replica version was reportedly a "marvel of engineering" as it consists of flesh made from silicone and a receptacle from 3D-printed plastic.
What the CREATE engineers have achieved was no small feat, as they have designed and built a fake raspberry that can "tell" the robot how much pressure is being applied to avoid damaging the fruit during harvest.
PhD student Kai Junge noted that through their sensorized raspberry and a machine learning program, they could teach a robot to apply just the right amount of force.
Junge, who is involved in the study, said the most challenging part was training the robot to loosen its grip once the raspberry detaches from the receptacle so that the fruit does not get squashed.
Considering that the robots are typically conventional, the released statement noted that this process was hard to achieve.
Read Also: Robots with Soft Grip? NUS Researchers Debut Machine with Delicate Fingers for Soft Materials
Ongoing Tests for the Fruit-Picking Robot
The engineers described the development as incredibly challenging since they have been using a simple feedback system. The harvesting robot currently consists of two 3D-printed fingers, acting as a gripper, covered with a thin silicone layer attached to a robotic arm.
PopSci reported that Hughes said the engineers' next step is to design and build more complex controllers so robots can pick raspberries on a larger scale without damaging or crushing them.
The engineers will first focus on developing a camera system that will allow robots to "feel" and "see" the raspberries. Hughes noted that this system would also be used to pick other soft fruits too.
"We'd also like to develop technology for other soft fruit and apply this physical-twin concept to more complicated tasks like other berries, tomatoes, apricots or grapes," Hughes said.
Based on a small test with real raspberries, the robot created by CREATE engineers harvested 60% of the fruits without damaging them. However, that's pretty low compared to the average 90% from human harvesters. Thus, they made a fake raspberry that could help the robot improve.
The CREATE engineers plan to test their robots in the summer raspberry-picking fields in order for them to see how well the machines perform. If this technology works as planned, the team reportedly wants to expand its fake fruit repertoire to cover apricots, tomatoes, and other berries.
Related Article: Cambridge Consultants Develops Robot That Can Pick And Sort Out Fruits
