“The produce industry is on the cusp of a major revolution in robotics. Labor-related issues are going to be addressed through technology, so why can’t we be part of the solution?” Gary Wishnatzki, Co-Founder of Harvest CROO Robotics and Owner of Wish Farms, asks me as we touch on the challenges and hurdles that face growers across the U.S.
In reality, Gary tells me that the demand is so great for a solution to alleviate labor shortages and create more cost effective in-field and harvesting technologies, that it isn’t just an East Coast or West Coast problem - it’s worldwide.
“Our goal is to be able to pick 95 percent of the fruit off of any plant.” - Gary Wishnatzki
In Gary’s eyes, the most challenging part is getting growers to think about what is possible. So, when Gary and Bob Pitzer, Chief Technology Officer and Co-Founder of Harvest CROO, decided that traditional means of labor needed to evolve, they set their sights on a strawberry bot that could harvest and pack strawberries with greater speed, accuracy, and efficiency.
“Robotics are changing the game. Our automated strawberry picker will be able to work at least 20 hours per day, and that includes weekends,” Gary laughs. “Our goal is to be able to pick 95 percent of the fruit off of any plant.”
The new Harvest CROO technology will allow growers to avoid picking during the hottest part of the day when berries bruise the easiest. In this sense it will take less time and energy to cool the fruit by picking in the evening hours. Also, the strawberry bot will allow for better utilization of cooling facilities and increase throughput by spreading out the load of warm incoming fruit over longer harvest hours.
Gary also tells me that shippers could add acreage without having to purchase additional cooling resources. Since the volume will be more spread out, there is also the potential to double the capacity that coolers can accommodate.
So, how does this strawberry harvester work? The patented technology revolves around the concept of the Pitzer Picking Wheel, named after Gary’s partner, Bob Pitzer. This wheel utilizes “conservation of motion principles” with robotic picking heads that can achieve 360 degrees of rotation and will decrease the amount of movement the robot has to accomplish to complete the picking action. A series of claws on the wheel picks the berries which are then transferred to a packing region of the harvester, where they will be inspected and packed into consumer units.
“What we hope to accomplish is to lower harvest costs by increasing the speed and duration at which it can pick and pack berries in the field."
To support the ever-evolving prototype, Gary and Bob have created a logo to accompany the new harvest technology and have appropriately named it Harv.
This picker also uses Stereo Vision with two cameras mounted on the harvester, with the lead camera situated to identify the berry’s color, mass, and size to decide whether or not it should be harvested. The second camera uses triangulation to pinpoint the berry for the claw to gently pick and place in the packaging.
“What we hope to accomplish is to lower harvest costs by increasing the speed and duration at which it can pick and pack berries in the field,” Gary explains as we continue to discuss the harvester’s far-reaching benefits for growers.
Weighing packages will be a huge savings to growers as well. Typically packed until visually full, over-packs can be up to 10 percent or more than their goal weight. Larger packs are estimated to be over-packed by as much as 20-30 percent, Gary shares. On the opposite side of the spectrum, this technology can also eliminate rejections due to being short weight.
Another possibility? Growers can potentially reduce their plastic usage by over 30 percent by using a film lid versus a clamshell pack, which the program plans to provide.
In regards to precision agriculture, the machine will be able to scout as it travels through the field, taking images of plants which will then be aligned with a database of hundreds of images that can provide early warnings of things like pest presence or plant damage.
“Once we get this project streamlined, it will open the doors for other opportunities,” Gary says. “There are automated lettuce harvesters, strides in pepper harvesting… This design could potentially be applied to other produce categories.”
At present, the harvester is only a solution for strawberries, but Harvest CROO’s team of engineers that Gary and Bob have been assembling will be looking to branch out into other categories in the future.
“We have a game plan that ensures that this prototype will continue to evolve. Technological advancements have allowed the harvester to progress in leaps and bounds, even in just the relatively short timeline that this project has been off the ground,” Gary adds.
Some of the robot’s recent developments include new designs for the picking wheel and leaf manipulator in order to improve accuracy and efficiency. With new and improved software, embedded electronics, and new motors, Harvest CROO is hoping to accelerate the picking process and pick a plant in under eight seconds. And the team is shooting to get faster than that. The inspection process of the plant takes a mere two seconds of that eight-second window. Also this winter, Harvest CROO is working on a new GPS program on a mobile platform for the production model.
“As we test this prototype out with strawberry growers, we are also making it our goal to try and adapt the machine to as many different varieties as we can,” Gary says.
The team has also improved on the ability to align the tractor to perform repeatable actions within a half of an inch.
“We have 16 picking wheels on one platform working in concert, and that is the goal,” Gary says. “Everything has to work in concert.”
All these prototypes are working towards a totally autonomous machine moving through the field. It is an exciting time.
“This whole entire project, while spearheaded by Bob and I, is truly a giant collaboration between the best minds in harvest technology,” Gary says.
Take Dr. Rex Lee of Pyramid Imaging, who has been working with Harvest CROO on the imaging platform and helping to develop the vision for the program moving forward. Dr. Lee has not only brought the latest and greatest in machine vision technologies to the project but the kind of like-minded, advanced thinking that is driving the bot’s progress.
“If a solution requires laws of physics to be broken, then it is impossible. Otherwise it's just an engineering problem and nothing is impossible.” This is one of Gary’s favorite quotes from Dr. Lee on vision, and, may I say, quite an inspiring one.
Bob’s background has also helped fuel the collaboration through multiple industries. "My involvement in the early Robot Combat television shows has helped me connect with many people in the robotics community,” Bob tells me.
Bob’s experience includes time in the Navy as a tech on a nuclear sub as well as attending the University of Florida, where he started multiple student robotics groups, and graduated with a mechanical engineering degree. Out of college, Bob was recruited to work for Intel in Arizona as an automation design specialist and later started his business, 4FX Design, which has worked with many companies to develop new and exciting technologies.
"There are many opportunities to enhance this prototype and it all depends on the support of many.” - Bob Pitzer
These experiences put Bob in a place where he spent years developing the technology to run the FIRST Robotics STEM education events which are developing the next generation of highly qualified engineers for the future while providing some of the brain power that now pushes Harvest CROO forward. Some of those relationships are part of the company’s current vendor list also. The man has been busy.
“Collaborating on this effort with people that I have a history with, gives me confidence,” Bob tells me. “An example is Velodyne's founder, Dave Hall. Velodyne is a leader in LIDAR technology used on autonomous vehicles and they are very excited about the possibilities represented by the agricultural industries. There are many opportunities to enhance this prototype and it all depends on the support of many.”
The Harvest CROO team recently tested the prototype this fall in Santa Maria, California, with a grower affiliated with California Giant Berry Farms. The initial debut has been great so far and the number of investors has grown exponentially since Gary and Bob began sharing the design with fellow strawberry growers in the industry. As the team moves the strawberry bot into the next phase of development, they are looking to continue perfecting the technology using varied prototypes to prove out different aspects of the project.
“There has been a tremendous reception for our strawberry harvester among some of the largest strawberry growers in the U.S.,” Gary says.
Harvest CROO has opened a funding round for those in the industry that want to invest in the program and has also established an early adopter program which gives investors priority during the initial phases of the launch as well as a nine percent discount for the first three years.
"Harvest CROO Robotics will be in a position to become the leader in ag robotics with the successful deployment of our picking robots. That is the vision,” Bob says.
At Wish Farms and Harvest CROO Robotics, innovation in agriculture doesn’t hope to replace the human element. But, if there are berries to be picked, you can be sure that this team will continue to find new ways to press on and up.