“This is just the first step of our automation journey,” explains Jacco Blok, Project Engineer at TOP The Onion Group.
For companies wishing to remain competitive, meet decarbonization goals, and face the challenge of a tight labor market, automation is essential. The TOP team decided to move towards automation on a timeline that works for the company as a whole before being pushed by these external factors.
When a manual process is working well, why introduce automation? You might argue, “If it ain’t broke, don’t fix it,” but there are compelling reasons to introduce automated vehicles before the manual process breaks down:
TOP The Onion Group expanded automation beyond its existing palletizer systems by introducing three Rhinos – heavy counterbalanced AGVs by KV-Techniek. These automated guided vehicles have taken on heavy, repetitive work and run up to 200 missions a day alongside human colleagues.
Work at TOP Onion's processing plant in 's-Gravenpolder is highly seasonal. Its workforce fluctuates in line with its orders, with between 60 and 120 people working at the plant at any one time. Automation takes some pressure off hiring and managing such a large team.
February is usually the busiest month of the year, as growers place orders for onion sets (immature bulbs for growers), in addition to ongoing retail orders for supermarkets and wholesalers. In February 2024, for example, all three Rhino AGVs ran a combined total of over 5,000 missions. In the off-season, one AGV might remain parked. However, unlike human workers, the vehicles can be brought back into service at the push of a button.
The Rhinos work solely end-of-line tasks, moving boxes of onions from the output conveyors of the palletizers to outbound staging (before being shipped out in trucks). From the staging area, the goods are moved onto trucks by a manual forklift.
What makes a Rhino a rhino? They’re big and fast.
The Rhino by KV-Techniek is a counterbalanced forked AGV weighing in at 4.4 tons. It can lift payloads of two tons up to 1,400 mm in height and moves at up to two meters per second.
The Rhino appealed to the TOP team in part because of its side shift function, which allows the forks to move laterally left and right. Since the pallets in the 's-Gravenpolder facility are narrower than the Rhino itself, this side shift function is essential to allow loads to be placed right next to each other without the risk of damaging the produce. The Rhino’s sensors ensure that while the pallets are packed so tightly and accurately that they touch, there’s no damage to the veggies.
TOP’s warehouse manager handles the specific orders of different wholesalers and retailers, determining for example in which outbound lanes different palleted orders should be placed. Since orders can arrive on EU or US-style pallets, the latter required a ‘counterbalanced’ AGV.
The AGVs work a loop in the outbound halls, picking up loads from the output conveyors of one of nine palletizers, depositing these pallets in one of several outbound staging lanes ready to be loaded manually onto trucks, before finally heading back to their charging stations. These transports are relatively short; an average loop might cover just 100 meters.
The Rhino’s job is to create a buffer or backlog of pallets ready for loading. Creating this buffer is steady, precise and repeatable work. The TOP team’s goal is to have all orders palletized and placed in the docking lanes in the right order for optimized loading well before the trucks arrive. “When doing this job manually, often someone places a pallet into the wrong lane. Many times in the past people have mixed up order numbers - like 36896 and 36869 for example. The AGV’s don’t make those kinds of mistakes,” says Blok. "Now, if something goes wrong it's likely to be because of the palletizers, or the connection between them and ANT server, not the AGVs themselves."
| Location | TOP The Onions Group's onion and onion sets processing & distribution center in 's-Gravenpolder, The Netherlands |
|---|---|
| Vehicles | 3 Rhino AGVs by KV-Techniek |
| Automation | BlueBotics ANT lite+ (vehicle navigation & localization) & ANT server (vehicle & fleet management software) |
| WMS | Custom web-based app by KV-Techniek |
| Installation Date | December 2021 |
| Sector(s) | FMCG, food processing |
The outbound workflow at TOP The Onion Group 's-Gravenpolder facility is spread across two halls, allowing the team to expand and contract as seasonal orders demand.
The main hall contains a row of eight palletizers, and docking lanes facing seven loading docks. In the second hall are two further palletizers and two docks. The AGVs’ chargers are located near the corridor linking the two halls.
The site’s full order workflow is concerned with the movement of goods from palletizer to outbound staging. Briefly:
In effect, the AGVs take on the grunt work of moving pallets from palletizer to docking lanes, creating a buffer of orders arranged for optimal loading. This dull, repetitive work is perfectly suited to mobile robots.
In TOP The Onion Group’s case, the AGVs work in tandem with manual vehicles, supplementing their efforts. The more complex truck loading task is still completed by a human worker who has been with the company for decades.
“All of our staff are getting used to working alongside automated vehicles. And we’ve all noticed the difference in accuracy between a person and a robot,” says Blok. "Unfortunately, we do still have to deal with damaged goods. But most of the time a person is to blame for that. For example, when a bad pallet breaks as the AGV picks it up and starts driving - that was less of a problem with manual forklifts, as the driver would notice the issue right away. However, we had other problems with manual forklifts, and with AGVs the pallets are placed better, with less damage overall."
Block storage environments (where items are stacked and stored without shelving or racking) are by definition dynamic environments. They can change from hour to hour as pallets are deposited or shipped out.
In such a block storage environment, where pallet numbers and positions are constantly changing, accurate autonomous navigation can be a significant challenge as the number of visible 'permanent features' - required to calculate a vehicle's position - is sometimes limited.
The ANT navigation technology inside TOP’s Rhino vehicles has several key features that makes accurate, repeatable navigation possible in such a tricky block storage environment:
‘ANT driven’ vehicles, such as the Rhino, localize themselves by 'looking at' the environment with their LiDAR laser scanners in order to identify permanent features such as walls, pillars and, if required, small reflectors. This process is called 'feature matching' because these permanent environmental features are matched to the vehicles' internal, 2D map in order for ANT to calculate a vehicle's position (localization) .
The system starts with a 'clean' map which excludes all non-permanent features. This means that while a robot will 'see' and be able to avoid people/pallets/other hazards on the shop floor, it will not mistake them for walls or other permanent features.
In a busy, ever-changing environment, ANT natural feature navigation is reliable and accurate.
> Read more about ANT natural feature navigation.
Installing the first Rhino on-site, in December 2021, took just three days. It was important to Blok and the rest of the TOP team that the project would not impact the entire production process. With 100,000 pallets passing through the facility every year, minimizing downtime was vital.
Before this first installation, the KV-Techniek team took a few weeks to learn the site, discuss the clients’ exact needs and expectations, and to simulate the site’s AGV operation in ANT software, ensuring for example that the vehicles’ routes were as optimized as possible. This was easy to do thanks to ANT software’s sophisticated Simulator tool, which allowed the team to work with a digital twin of the space to plot the best routes without having to impact real-world operations.
In parallel, the BlueBotics software team developed a new tool to assist in the installation: ANT server’s Lane Manager. In large warehouses like TOP The Onion Group’s 's-Gravenpolder facility, outbound pallets are stored in lanes that are packed tightly together to optimize space. As a result, there is not enough space for two AGVs to pass in adjacent lanes. The Lane Manager tool solves this potential deadlock situation by allowing a user to specify which lane zones can only accept one vehicle at a time. It’s a simple change, but an effective one.
The Lane Manager tool was inspired by the TOP project and worked upon by BlueBotics and KV-Techniek, but is available for all ANT customers – and is already in use in other sites. “It can be set up in seconds and save a lot of issues,” explains Twardowski.
Following the vehicles’ installation and go-live, KV-Techniek’s team offered intensive aftercare over several months, to ensure the system was running smoothly, and build knowledge to prevent issues from occurring in the future.
Part of Blok’s job is to manage the AGVs and other automated machinery day in, day out. He is fully trained on the AGVs’ ANT navigation and fleet management software, meaning he can make configuration changes to the site’s AGV operation himself.
Since the installation of TOP’s AGVs in 2021, 2022 and 2023, the system has been running continuously, with no major changes required to the process, proof of KV-Techniek’s 'get it right first time' philosophy. As expected, the AGVs work exactly as programmed every time, with no mistakes or damaged goods.
And there’s room to grow. With three AGVs at the ready, the team has some extra capacity, ready for a bumper year. The project has been so successful, the TOP team plans to increase their automated vehicle fleet in the coming year.
“We as TOP The Onion Group felt the need to automate as it is more and more difficult to get good employees. Since the Rhinos came online, there have been no more mixed-up orders due to human error, and less damage to our conveyors and goods. The AGVs don't get stressed when busy - they keep going with a constant pace, while a person is more likely to make mistakes under pressure. These AGVs were a great first step in our automation journey,” concludes Blok.