Optimal control of AGVs and AMRs

Driverless transport vehicles move material from A to B. What sounds simple is actually quite complicated: The more vehicles are on the move on the shop floor, the greater the challenges become. What is needed is a smart control system.

Nowadays, every order is an individual order. In e-commerce, for example: each shopping basket is unique and made up of a wide variety of items that first have to get from large warehouses into the parcels. Hoover bags, reference books and a new pair of sweatpants have to be packed and shipped as quickly as possible.

Efficiently managed intralogistics are enormously important because customers would prefer to have their ordered goods in their hands immediately. An online shop can hardly afford a slow order processing that may take days - the competition is too brutal. After all, other providers deliver the very next day and thus also improve customer loyalty.

The configuration and assembly of a new car is also complex. Genuine one-offs with selected special equipment roll off the assembly line. What makes individualists particularly happy presents manufacturers and their intralogistics with special challenges: When assembling the dream cars, the many individual parts have to be in the right place at the right time. If they do not reach the production line in time, expensive delays are imminent.

EFFICIENT INTRALOGISTICS

In production halls and warehouses, automated guided vehicles (AGVs) move the required materials from one place to another. What arrives at the goods receiving area is first taken to the warehouse and stowed there. In the further course, the parts have to be transported, for example, to final assembly, intermediate storage or dispatch processing.

AGVs (Automated Guided Vehicle) and AMRs (Autonomous Mobile Robot) take care of the diverse transport orders and act autonomously. It is precisely here that processes must be carefully planned, efficiently optimized and finally executed. What matters here is holistic optimization - not step-by-step planning. The desired goal is ideal resource utilisation and maximum delivery reliability. The focus is on timely delivery of all required parts and materials.

Traffic jams, accidents or technical malfunctions interrupt the schedule and disrupt careful planning. Flexibility is therefore required - and above all: fast action. Humans can only do this to a limited extent; the complexity and the amount of data that must be incorporated into sensible planning are too great. One-off or infrequently adjusted top-down planning is not enough because the modern shop floor is highly dynamic.

NON-STOP OPTIMIZATION IN REAL TIME

Continuous optimization in real time is the order of the day: control software must react to events without any delay and reschedule quickly, over and over again. A vehicle suddenly breaks down? Another one has to step in immediately and take over the job! The SYNAOS Intralogistics Management Platform reschedules 250,000 times per second and always looks for the very best solution so that a disruption does not cause a long-lasting delay. Powerful algorithms distribute tasks, determine routes and continuously check whether routes need to be rescheduled.

Such process optimisation quickly becomes highly complex because numerous interdependencies and chains have to be taken into account in the planning. Every disruption can have a knock-on effect - like the famous butterfly effect: a small flap of wings can possibly trigger a tornado on the other side of the world.

In intralogistics, any small change and delay can cause unpredictable problems, which is why control software must constantly recheck and react to events. This is exactly what adaptive algorithms do, which continuously process data in real time and optimize processes on the basis of this information. The more information, the better. Complexity is a real advantage in this case!

HOLISTIC INTERACTION

In addition, there are other circumstances and conditions that need to be taken into account during planning, such as different order characteristics: Which container type is in use? What are the AGV characteristics? Traffic management, including route optimisation, also plays an important role. Ideally, the AGVs and AMRs take the optimal route in each case (which does not necessarily have to be the shortest). What is important is the holistic interaction and optimisation as a whole.

The planning also includes the infrastructure, specifically the traffic lights, gates and systems in the halls, which also have to be controlled in coordination with the rest of the system. Energy management, i.e. the power supply for the vehicles, also plays an important role. As long as they are filling up with electricity at the charging station, they are not available for orders. A well-managed energy budget leads to punctually completed orders and keeps the entire fleet fit.

In addition, various map information flows into the control system, i.e. the layout and structure of the warehouses and production halls. Data such as speed limits, one-way streets, temporary closures due to construction work and so on are also included. So there is a huge amount of data that can flow into an optimisation and cause highly complex processes.

BIG CHALLENGES

Despite advancing digitalisation and automation, fully autonomous fleets are rare. In reality, manual participants are additionally on the move in the same hall, i.e. human employees who transport goods with a forklift truck, for example. Such a mixture of manually operated vehicles and fully automated transport systems leads to unpredictable dynamics and turbulence. For the control system, this results in another challenge.

There are also certain problems to overcome when exchanging data: Each AGV and AMR manufacturer has come up with its own solutions for communicating with the vehicles and has developed proprietary software. However, this heterogeneity in communication has major disadvantages because it prevents uniform control. It also makes it impossible to operate mixed fleets with AGVs from different manufacturers.

For users, this leads to a so-called vendor lock-in: once they have decided on an AGV brand, they are tied to this one manufacturer and dependent on it in the long term. It becomes annoying when a competitor has a much better and at the same time cheaper solution in its portfolio, but this then does not work with the vehicles already available. Moreover, not every manufacturer offers AGVs and AMRs with all transport and handling capabilities. This limits the application scenarios for customers. However, there is already a solution for these problems: the VDA 5050 interface.