Control AGVs optimally using VDA 5050
Controlling automated guided vehicles is complex – and this is particularly true for holistic process optimization. A standardized interface now simplifies communication and enables the operation of heterogeneous fleets with vehicles from different manufacturers. The possibilities of the VDA 5050 standard are manifold.
Nowadays, every order is a custom order. For example, in e-commerce: Every shopping cart is unique and may contain a wide variety of products that must first be transported from large warehouses into their packaging. A vacuum bag, textbook and new sweatpants must all be packaged and shipped as rapidly as possible. Efficient intralogistics management is extremely important: after all, when customers order products, they want to receive them as quickly as possible. No online shop can survive now with sluggish order handling that takes a matter of days – the competition is simply too tough. Other providers can offer next-day delivery, and they strengthen customer loyalty by doing so.
The configuration and assembly of a new car is also highly complex. The vehicles coming off the conveyor belt these days are not just dull standard models, but unique items with selected special equipment. These developments might make individualists happy, but they pose particular challenges for manufacturers and their intralogistics systems: ultimately, assembling a customer’s dream car requires many individual parts to be in the right place at the right time. If the panoramic sunroof or powerful LED headlights don’t arrive at the assembly line on time, this could result in expensive delays. Production and intralogistics have to efficiently handle many orders in sequence. For this process, it is essential to ensure a continuous flow of goods and internally coordinated transport processes.
In production halls and warehouses, Automated Guided Vehicles (AGV) move the required materials from one place to another. Incoming goods are first brought to the warehouse and stored there. In the next step, the parts must be transported for final assembly, interim storage or shipment handling. Automated Guided Vehicles take care of these various transport tasks and operate autonomously. Precisely in this context, processes must be carefully planned, efficiently optimized and finally carried out. The method required here is holistic optimization – not just step-by-step planning. The desired goal is to achieve ideal utilization of resources and maximum delivery performance. The key is on-time delivery of all required parts and materials.
Accidents, bottlenecks, breakdowns
Bottlenecks, accidents or technical breakdowns interrupt the process and rapidly thwart even the most cautious of plans. What is therefore required above all is flexibility and a rapid response. Human beings can only manage this to a limited extent – the complexity and volume of data involved in sensible planning is simply too large. Top-down planning that is only adjusted once or infrequently will not be sufficient, since modern shop floors are highly dynamic.
The shop floor is always busy. Automated guided vehicles transport materials from A to B – ideally, they will always take the optimal route. If problems arise, the control software has to react immediately and reschedule the routes.
A modern solution requires continuous optimization in real time: A control software must react to events without any delay and reschedule rapidly, again and again. What if a vehicle suddenly breaks down? Then a replacement has to jump in and take over the task. This is exactly where SYNA.OS LOGISTICS comes in: The AI based software makes new plans 250,000 times a second and continuously looks for the optimal solution to prevent a breakdown from causing serious delays. Algorithms distribute tasks, define routes and continuously review whether routes have to be rescheduled.
This kind of process optimization quickly achieves a high degree of complexity because countless dependencies and interconnections must be considered during planning. For example, the coordination of processes with filled containers and empty containers: Container spaces are generally occupied by full containers; then parts are removed until the container is empty and needs to be replaced. The empty container needs to be transported away before a new container can be positioned. If this is not carried out in time, even if a vehicle is ready with a new container and the required parts, it will have to wait unnecessarily or even block the retrieval of the empty container. This can trigger a sequential build-up similar to the famous butterfly effect: The smallest flap of a wing could trigger a tornado somewhere on the other side of the world. Any modification or delay in intralogistics, no matter how small, can cause unforeseeable problems. As a result, control software has to continuously review the situation and react to events. This is exactly the job of modern algorithms, which process data in real time without interruptions and optimize processes based on the data. The more information, the better. In this case, complexity is a definite advantage!
Holistic interplay
Other circumstances and conditions exist which must also be considered during planning, for example different order characteristics: Which type of container is used? Which AGV characteristics are involved? Traffic management including route optimization also plays a crucial role. The halls are always very busy, lots of vehicles are moving around and transporting material from A to B. Ideally, AGVs always take the optimal route (which does not necessarily have to be the shortest). The decisive factor is the holistic interplay of elements and overall optimization. Planning also considers the infrastructure: specifically, this means the stoplights, gates and systems in the halls that must also be controlled in coordination with the rest of the system.
Vehicles can’t rush around the halls forever: at some point, their battery runs out and the vehicles have to head to the charging station for a period of time. Two levels must be considered here: On the first level, each AGV has a specified threshold value. If the battery percentage falls below this value, a charging job is triggered. The vehicle is no longer available to accept jobs for a certain period of time and heads to the charging base before the battery is fully drained. The second level is the overall energy level of the fleet, i.e., the charging status of all vehicles that transport goods through the halls. It is important to charge the vehicles proactively to maintain continuous long-term operation. Otherwise, all the vehicles might determine at some point that they are about to run out of power. They would all make their way to the charging station at the same time, and the hall would suddenly go quiet. Instead, the energy management system must ensure that vehicles are always available with a sufficient charge for impending transport tasks while other vehicles are proactively recharged. Proper energy management means that orders are completed on time and the entire fleet is kept in good shape.
The control process also considers various map data, for example the layout and design of the warehouses and production halls. Data such as speed limits, one-way streets, temporary closures due to construction work and the like are also included. Massive volumes of data might be incorporated into the optimization process, resulting in highly complex processes.
Major challenges
Despite the progress of digitalization and automation, fully autonomous fleets are rare to see. In reality, manual actors are also moving around the same halls, for example human employees who transport goods on a forklift. This mixture of manually operated vehicles and fully automated transport systems leads to an unpredictably dynamic and convoluted process. This poses a further challenge for the control process.
In the past, a great deal of emphasis was placed on good hardware. Customers do not necessarily want to purchase state-of-the-art robots with fantastic capabilities, but rather need a reliable solution for optimally efficient intralogistics. Now as before, hardware manufacturers do not always have sufficient capacities to extensively develop their software further and implement new technologies that may become available. The scalability of software and control solutions is limited due to the resulting technical capacities. A strict separation of hardware and software can resolve this problem. This is why SYNAOS consciously decided to focus solely on software and to cooperate with strong hardware partners.
Still, certain problems need to be overcome when exchanging data: Each AGV manufacturer has conceived of its own solutions for communication with the vehicles and developed proprietary software. But this diversity in communication methods has major disadvantages, since it prevents unified control. It is also impossible to operate mixed fleets with AGVs from different manufacturers. As a result, users are subject to what is known as vendor lock-in: Once they decide on an AGV brand, they are stuck with that one manufacturer and dependent on them for the long term. The situation becomes frustrating if a competitor has a much better and more affordable solution in their portfolio, but it is not able to work together with the existing vehicles in the fleet. What’s more, not every manufacturer offers AGVs and AMRs with all capacities for transport and maneuvering. This restricts the possible applications available for customers. Nevertheless, a solution to this problem already exists: VDA 5050.
VDA 5050 standardizes the exchange of data
Standardization of communication between hardware (the AGVs) and software (control system) enables the operation of manufacturer-independent fleets. Increased heterogeneity offers many advantages for the user, for example by enabling financial savings and optimal flexibility. Simplification and unification offer additional advantages by increasing efficiency and quality.
The VDA 5050 interface makes it possible to control transport vehicles independently of their manufacturer. As a result, fleets can be extremely heterogeneous. Users also benefit financially from this diversity.
The VDA 5050 specification can be described as a common approach for standardizing communication. This project was initiated by users, specifically by the German Association of the Automotive Industry (VDA). Installations have continued to grow, but no solution on the market was able to cover all the requirements. The new interface makes Plug & Play a reality even for intralogistics and production: A new vehicle should be hooked into an existing system quickly and take on transport jobs rapidly.
In many areas of consumer electronics, this has already been the case for a long time: Users can easily plug a new printer into their PC via USB. They do not have to worry about the cumbersome process of driver installation and configuration because the operating system takes care of it. New AGVs and other vehicles and robots that are integrated into an existing fleet need to be up and running as quickly as possible. This makes it possible to create a heterogeneous fleet in a short time and expand the fleet easily as needed. A national standard – which will be recognized internationally at some point – should also strengthen German manufacturers and leave them in a better position globally. After all, the aim is to rapidly commission AGVs from German manufacturers even in China.
Uniform communication
The VDA 5050 is a standardized communication interface between the automated vehicles and the backend software, that is, the master control system. The standard provides a unified data language and defines specifically how jobs, map information, the status of AGVs and errors are communicated. In addition, the VDA 5050 even prescribes technical implementation: For example, communication is defined via the robust MQTT standard (Message Queuing Telemetry Transport). This is an open network protocol that enables reliable communication at a low bandwidth despite large delays. MQTT is an asynchronous messaging format and uses data formats such as JSON (JavaScript Object Notation). MQTT is used for applications such as sensors, cell phones, embedded systems or vehicles. All these devices speak a common language with MQTT.
A MQTT server acts as a central broker that distributes the collected data. The broker has a complex situational awareness and acts as a strong communication partner. The individual participants do not communicate with one another. Instead, they each communicate exclusively through the broker, which receives, sends and manages messages. Each actor establishes its own connection with the broker and receives the messages with topics that are relevant for them. Clients can subscribe to these topics; the broker forwards received messages to the relevant subscribers. There are advantages to this type of communication, for example for small devices with insufficient capacity for continuous data queries: Instead of the device, the broker delivers the required information to other devices or connected software. For this reason, MQTT is used in areas such as smart homes and IoT devices in general (Internet of Things).
Apart from automotive manufacturers, AGV manufacturers are also involved in developing the VDA 5050 standard, specifically the VDMA Materials Handling and Intralogistics trade association. Experts are developing the standard on a joint project team, including the Karlsruhe Institute of Technology (KIT) and SYNAOS.
VDA 5050 at SYNAOS
At SYNAOS, VDA 5050 onboarding is a crucial element of a typical project process. This structured process allows hardware manufacturers to implement the VDA 5050 for an AGV with a high degree of efficiency and quality. SYNAOS also develops validators to analyze and check whether communication is actually VDA 5050-compliant. SYNAOS has built up a wide network of AGV partners for improved synergy and distribution of VDA 5050.
The new interface passed its first public live test in March 2021 at the AGV Mesh-Up. This multi-day test was part of the IFOY TEST CAMP in Dortmund. There, a central control system controlled the automated vehicles of six different vehicles and guided them reliably across an activity space. This supplied important evidence of the capacity for interoperability. At SYNAOS, we began successfully controlling automated vehicles from the cloud using VDA 5050 on our own premises in late 2019.
The number of intralogistics providers who use the VDA 5050 standard has already grown considerably. In the future, stakeholders aim to further pursue and reinforce the open-source and community approach of VDA 5050. For example, the source code will be published openly on the development platform GitHub to promote rapid further development. Another important goal is for the VDA 5050 standard to be incorporated into the DIN and ISO standards and thus find international recognition. The potential of this new standard is far from exhausted, and the prospects are highly promising.
For more information visit our website www.synaos.com or contact us at info@synaos.com.
About the author
Dr. Lennart Bochmann
Lennart is Co-founder and CPO of SYNAOS. With years of experience in the manufacturing
and logistics industry he is striving for bringing market-ready and customer-centric software
products to life that leverage the full potential of digitization in factories and warehouses.
Contact: info@synaos.com
