Unmanned Systems Technology 042 | Mayflower Autonomous Ship | Embedded Computing | ElevonX Sierra VTOL | UUVs insight | Flygas Engineering GAS418S | Ocean Business 2021 report | Electric motors | Priva Kompano

103 vehicle concept for harvesting jobs, so it’s easier to get something like that onto the market in the short term than autonomous harvesters or similarly complicated solutions.” The value of an unmanned vehicle capable of cutting away tomato plants’ leaves has therefore driven Netherlands-headquartered multinational Priva to develop its Kompano autonomous de-leafing robot, which is now commercially available. The company is focusing much of its efforts not only on r&d for autonomous agricultural solutions, but also ensuring sufficient competency and infrastructure on the part of its customers for successful operations with the Kompanos. There are several stages and facets to the correct use of this UGV, and its design and operation have come as a result of many years of work by both Priva and its various partners in farming, technology and other key areas of expertise. History of the Kompano The Kompano is a 191 cm-long, 88 cm- wide, 180 cm-tall four-wheeled vehicle, weighing around 425 kg and running on a 5 kWh lithium battery. It has a robotic cutter arm that extends up to 125 cm. It is built using a combination of metal and plastic parts; the chassis is largely metal, and plastic makes up most of the body. Its arm extends from the centre, the battery pack sits in the rear, and at its four corners are LED lights that function as indicators. Coen says, “The idea and project for conceiving an autonomous de-leafing solution has been around in Priva since the early 2000s. In fact I was still in university when they first started.” In 2001, some growers approached Priva to discuss the issues they were having with finding labour for de-leafing work, and suggested that this could be an easy task (relative to harvesting) to develop an autonomous solution for. “It’s not primarily about engineering – that’s something to remember about agricultural robotics, and it’s a mistake that a lot of companies are making,” Coen comments. “It’s about understanding the plants, about having the agricultural knowledge to make the robot work correctly, because there are key requirements as to what the robot must not do. “For instance, if you want to automate harvesting, first you can’t allow crops to become bruised. Then you need to do different jobs at specific times – if you leave a job for a few days after the optimal time, you get a sub-optimal crop. “There’s no point in a robot that gives you sub-optimal crops, the whole point of investing in an autonomous solution is for it to be as good or better than a human worker. And there are many more issues of logistical, mechanical and visual difficulty. “When it comes to de-leafing specifically, your ‘cut’ needs to be performed very nicely, and that’s about it in terms of how complicated the job is. What happens to the leaf afterwards isn’t important, and if you miss one or two, that’s fine, it can be left there until a day or two later. The bar for difficulty is certainly lower for leafing than for harvesting.” Priva subsequently became interested in moving into the unmanned space. The company initially worked on concepts, which took place over several years. Coen points out that the early 2000s were not flush with UGV manufacturers to take inspiration and guidance from, so concept design was anything but straightforward. By 2008, the company had gained its first patent for an end-effector arm to remove leaves from stalks, and following that it continued working on the prototype of the UGV. The completed prototype was first showcased in 2016, winning an innovation award at the GreenTech horticulture expo in Amsterdam. “Shortly after that however, we had some issues with production partners, so we took a slight step back to do some design iterations of the prototype, which led to the new model we showcased Priva Kompano | In operation Unmanned Systems Technology | February/March 2022 At the top of the UGV is an end-effector for trimming leaves from tomato plants; computer vision is key to selecting which leaves to cut and the order thereof