When developing a new Autonomous Mobile Robot (AMR) for intralogistics, deriving key lessons from the real-world use cases from the very beginning of the journey is crucial. This is what arculus did with arculee M development. In this blog post, Carlo Fitz, our Managing Director; Romano Wolf, our Product Lead for Robotics; and Iuri Ferreira, our Release Coordinator, discuss the what and how of the development process.
New AMR for Efficient Intralogistics: Goals and Strategy
Let's start with the basics. What was the goal and the development strategy for the arculee M?
Carlo: We have been part of the Jungheinrich since 2021. We began this partnership with our first Autonomous Mobile Robot (AMR), the arculee S. Since then, we have gained extensive field experience with the arculee S. We received valuable feedback from Jungheinrich's sales organisation and product and portfolio management. Based on this feedback, we developed a new product increment: the arculee M.
Romano: At the beginning of our partnership, we analysed Jungheinrich's installed base of pallet transport systems. This analysis focused on developing the arculee M specifically for pallet transportation. By examining this extensive database of installed projects, we gained valuable insights into the necessary criteria for various use cases, such as dimensions, load capacities, and other specifications.
With this information, we could precisely tailor the arculee M to meet the needs of our addressable market from Jungheinrich's perspective. This allowed us to identify the optimal specifications for our new robot and address gaps that arculee S could not fill. As a result, we expanded our market share and met the customer needs even better than before. This was our primary goal from the beginning of the development.
Can you introduce the arculee M as a system? Please explain how it's designed to play a key role in intralogistics applications.
Romano: We need to understand the arculee M as a system from two perspectives. On the one hand, there is robotics, which includes peripherals like stations and backpacks necessary for pallet transportation. On the other hand, we have fleet management, which is essential for integrating with warehouse management systems, also known as host systems. For this, we utilise a software layer from Jungheinrich, the logistics interface, which acts as an abstraction layer for our fleet management. Thus, the arculee M system encompasses all robotics components, the fleet management and the logistics interface.
Carlo: From the customer’s perspective, you get transportation from point A to point B, where the robot performs the physical tasks. However, when replacing manual transportation with automation, it's important to consider various customer interfaces. The host system manages these interfaces, determining where transportation is needed. This makes the arculee M part of a larger system with numerous software components.
What are some common pain points in intralogistics that the arculee M aims to solve? Can you give examples of how it tackles these issues?
Carlo: Logistics is a cost factor but also a crucial enabler. While transporting goods from A to B doesn't seem to add much value, having them in the right place at the right time is essential. This is where the arculee M comes in. This new AMR enables efficient transportation while minimising costs. It performs reliably 24/7, making it an effective substitute for a cost-intensive intralogistics system.
Moreover, transportation systems like conveyors were very reliable and fast in the past. However, they were also inflexible. Changing the configuration, such as the source or sink, was costly and required disassembly, reassembly, and significant infrastructure adjustments. The arculee M system, on the other hand, offers much greater flexibility. The initial setup for a project can evolve without incurring high costs.
Romano: There are several reasons why our customers want to automate. One major reason is cost. Over time, they aim to amortise their investments in optimisation, achieve scaling effects, and become more cost-effective. While cost was the primary driver in the past, recent changes in the labour market have shifted priorities. Nowadays, labour is hard to find. So many customers are driven to automate because they cannot secure the necessary workforce for manual tasks. Companies must invest in optimisation to maintain their business operations and remain productive.
The arculee M Use Cases
What are the primary use cases for which the arculee M was developed?
Iuri: The arculee M tries to improve the solutions for underload carrier transport, meaning the arculee M would drive underneath the load, lift it, and carry it around in a plant. Its primary focus is the intralogistics, transporting goods from people to storage. This includes use cases such as narrow aisles with highly stacked items, in conjunction with other automation solutions that Jungheinrich also provides, to move goods around in parallel efficiently.
Romano: The arculee M is specifically designed for pallet transportation, particularly for Very Narrow Aisle (VNA) applications. These are high-rack storage solutions, and Jungheinrich is already a major player. We saw a significant opportunity to improve pallets' transport into these high-rack storage solutions.
One key advantage is our ability to connect two automated trucks from Jungheinrich and arculus, deploying them as a single, integrated system. However, this is just one example. Overall, the arculee M focuses on pallet transportation and can be used in various settings, such as warehouses and production facilities, where pallets must be moved from one location to another. The system efficiently picks up and shuffles pallets at designated stations.
How flexible is the new robot for specific client needs and requirements?
Carlo: The core idea was to balance these two ports. We created a physical platform with the arculee M, incorporating all the essential, complex functionalities such as localisation, navigation, and lifting activities. Based on this platform, we developed a backpack system that allows different backpacks for different load carrier types and makes the system flexible for various applications.
Additionally, we redesigned the software layer to focus on easy deployment and system robustness. This included creating a layout that minimises the number of robots used, increases throughput, and minimises deadlocks.
Romano: When we look at the market, we see a huge variety of pallets and other load carriers, and from the beginning, we aimed to stay flexible. So, how could we ensure this? We didn't want to develop a different product for every load carrier type. The idea behind the arculee M was to keep all the robotic complexity within the platform and the robot itself. Then, we created a mechanical and electrical interface for the backpack, which can be customised for different load carriers.
Iuri: This approach allows our software to remain mostly standard and basic, making testing and ensuring robustness and performance easier. At the same time, it provides the customer with the flexibility to customise the backpacks as needed, supported by the logistics interface.
The Development
Let's talk development. How are technical specifications and performance benchmarks determined for a new robot?
Iuri: When developing a new robot, the first step is to examine the market. We identify the market points we want to cover and define our target use case accordingly. We then analyse the market competition. With this information, we evaluate our technology stack to determine how we can leverage our existing capabilities to meet our use case requirements and what modifications are necessary. Navigating these modifications presents an interesting challenge. The more changes we introduce, the more complexity and risk we add to the development process, and the harder it becomes to meet deadlines.
At the end of the day, our goal is to create a high-performance robot that maximises throughput for the customer. So, there's always a fine line between further increasing robustness and stability while minimising risks and bringing a good product to market.
Romano: We call what Iuri just mentioned the platform approach. Our strategy involves developing products vertically, meaning we have access to all platform components ranging from the high-level software layer to the low-level software running on a dedicated microcontroller.
When we develop a new derivative, we first examine the use cases to identify relevant requirements and product specifications. Based on these insights, we can adapt and scale our platform as needed. This might involve adding peripherals or sensors or modifying the physical aspects of the platform. These changes could be in mechanics, electronics, or software.
The key advantage of this approach is that it is always based on our existing platform and expertise. This allows us to efficiently scale our platform to meet the new use case we want to target.
Carlo: The goal is to create a minimum viable product (MVP) with specific features that can be brought to market quickly. This will allow us to gather customer feedback and understand the market demands. From there, we will iterate on the product, refining it based on real-world data.
Initial ideas for a product are based on hypotheses. It's crucial to test these hypotheses early in the process to validate or correct our assumptions. This helps us determine if our initial market hypotheses are accurate. Through this iterative approach, we can evolve the MVP into a final product that effectively meets market needs.
The Evolution from S to M
Finally, let's talk about evolution. How does the arculee M signify an evolution from the arculee S regarding functionality and benefits?
Romano: For the arculee S, we initially targeted table transportation. At that time, we had major projects in warehouse applications, and transporting tables from point A to point B was a common task. This was the primary purpose of the arculee S. However, after Jungheinrich's acquisition, we discovered the field of pallet transportation. Our first project in pallet transportation was done using the arculee S, and it was successful.
However, despite this success, we encountered limitations as we brought the product to market, particularly in scaling up load capacity and height. When reviewing potential new projects, we noticed a gap in our portfolio—we couldn't handle higher load requirements. This realisation led to the development of the arculee M. It was designed to close this gap and accommodate higher load capacities and greater load heights, addressing the limitations of the arculee S for the pallet use case.
Iuri: The improvements don't stop there. As we discussed earlier, robotics is a dynamic, evolving field. Based on field feedback, we significantly improved the robot's serviceability. We focused on how easy it is to provide service and maintenance. The robot includes several critical safety features that must be tested annually. With this in mind, we made changes to simplify the service process for technicians, ultimately reducing costs and enhancing the overall product.
Romano: Specifically for the arculee M, we divided the product into several submodules, such as the lift, active lift, passive lift, electronics module, and peripherals. This modular approach makes the robot easier to assemble, significantly reducing production time. Additionally, it simplifies maintenance, making it easier to access, service, and replace components. Achieving these goals has been a significant success for us.
How has field experience from the arculee S shaped the development of the M model?
Carlo: The arculee M inherited many core ideas from the S. For example, we retained the central Robot Control Unit (RCU) in the arculee M, which we developed for the S. The RCU is a closed electronics unit that incorporates all the key electronic features needed to create the robot's autonomy layer. However, for the improvements, in addition to focusing on load capacities, we concentrated on enhancing localisation and navigation capabilities, putting significant effort into software development in these areas for the new robot.
Romano: Looking at the arculee S and its components, we learned from field experience which components worked well and which posed challenges due to supplier issues or performance. We applied these lessons to the arculee M. We kept or adapted the components that performed well while we upgraded components where we have seen potential for improvements. This included enhancements to actuators, sensors, and software.
This is all about how focusing on particular use cases improved the production of arculee M. Watch the complete interview here!