AMRs in intralogistics: from flow analysis to a reliable deployment

How can a manufacturing plant be modernised at a time of rising labour costs and growing pressure to meet deadlines? Automating the assembly line is important, but productivity depends on more than the assembly process itself. Even a highly efficient line will not deliver the expected results without well-planned intralogistics. Autonomous mobile robots can provide this essential support.

In today’s challenging economic environment, many manufacturers face rising labour costs, pressure to meet deadlines and maintain reliable supplies, as well as increasing production variability caused by short production runs and frequent product changes. Automated assembly lines are a natural development path for many plants because they improve productivity and process repeatability. However, the processes immediately upstream and downstream of the line are still often overlooked or treated as secondary.

This is a significant design and organisational mistake with direct business consequences. An assembly line may be designed to meet the required throughput, cycle time and process stability, but it cannot maintain these parameters without the timely delivery of components and auxiliary materials, the collection of finished products and empty packaging, and effective buffer management. Line performance is directly dependent on the efficiency of the intralogistics processes that supply and support it.

Unfortunately, management often recognises only after a new production line has been installed and commissioned that reliable, ongoing intralogistics support must also be organised and that the original assumptions regarding component flows are insufficient. The problem then has to be resolved immediately because the line must keep producing, meet its targets and generate revenue.

Why more operators and forklifts do not always solve the problem

The first instinct is often to increase staffing in intralogistics, sometimes alongside the purchase of additional forklift trucks or other transport equipment. Although this approach is intuitive and relatively quick to implement, it does not always produce the expected result. In practice, it can increase traffic, raise the risk of collisions, create bottlenecks and make the process even more dependent on operator availability.

When visiting manufacturing plants in Poland and elsewhere in Europe, we frequently see congested transport routes, pressure on employees responsible for material handling and difficulties during peak logistics periods. On the shop floor, instructions such as ‘Get this to line three quickly – they are waiting to start production’ are not uncommon. This is not a stable process; it is continuous reactive management and operational fire-fighting.

Although the production plan should support effective intralogistics planning, experience shows that manually managed internal transport is often too inflexible to cope with breakdowns, downtime, changes in order sequence, changeovers and sudden schedule adjustments. It is under precisely these conditions that intralogistics automation becomes particularly valuable.

One of the most flexible tools used in this area is the autonomous mobile robot, or AMR.

A brief technical overview of AMRs

AMRs are autonomous transport vehicles that move around a production facility without fixed tracks, inductive wires or magnetic strips. Depending on their design, application, payload and workload, they may operate for several hours or for most of a production shift. Many applications also use opportunity charging, allowing the robot to recharge its batteries during short breaks between missions.

There is no universal operating or charging time that applies to every device. These parameters depend on factors such as battery capacity, payload, travel speed, number of stops, docking method, ambient temperature and route characteristics.

An AMR is more than a device that moves material from point A to point B. A complete system includes the robot, its work module, docking points, charging stations and a fleet management system. Integrating these elements with MES, WMS or ERP systems enables transport missions to be generated and prioritised automatically, allows the system to respond to changes in the production plan and provides control over fleet utilisation.

AMR payloads and work modules

AMRs are available with a wide range of payload capacities, from tens or hundreds of kilograms to several tonnes. Specialist designs can also transport substantially heavier or oversized loads. Selecting the right device requires consideration not only of its nominal payload, but also of the position of the centre of gravity, transport dynamics, load-handling method, docking requirements and, crucially, the condition of the floor.

AMR advantages, transport missions and application flexibility

One of the main advantages of AMR technology is the ability to equip the robot with a custom-designed work module, often referred to as a top module or topper.

Examples include:

• an electric lifting module designed to work with dedicated transport racks,
• a roller conveyor module for collecting and transferring pallets or containers,
• a hook system for towing transport trolleys,
• a module that lifts an entire platform or rack,
• a module for transporting KLT containers, crates, racks or other load carriers,
• an industrial manipulator or robotic arm integrated with a vision system or another active system.

Combining an AMR platform with a robotic arm creates a mobile manipulator that can not only transport material, but also pick it up, place it, sort it and service selected workstations. Such a solution is sometimes described as a mobile robotic operator. However, using a collaborative robot does not automatically mean that the complete unit can operate without additional safeguards. The safety of the entire system must be assessed individually, taking into account the gripper, load, speed, reach and the way in which the system interacts with people.

The ability to equip a single AMR with different automation modules provides considerable flexibility in the tasks it can perform. In AMR terminology, transport tasks are usually referred to as missions. These may include delivering a component, collecting an empty container, transporting a finished product, replenishing a buffer, transferring a pallet or servicing a specific process point.

Fleet management and integration with MES, WMS and ERP

A central fleet management system integrated with MES, WMS or ERP, as well as with production lines, warehouses and peripheral equipment, creates a complete intralogistics solution. Its role extends beyond controlling robot traffic: it also queues tasks, assigns priorities, manages charging, prevents route conflicts and reports mission progress.

With properly designed integration, a change to a production order in the MES can automatically trigger a sequence of logistics operations: removing materials allocated to the previous order, supplying components for the new order, preparing buffers and delivering the correct containers to the line. This functionality does not arise simply from owning AMRs. It requires process logic to be defined in advance, reliable material identification, system integration and a correctly structured data model.

The facility map and process points can also be updated relatively quickly when a machine or workstation is relocated. However, every change should be verified in terms of safety, route availability, aisle widths, docking and its impact on other missions. Particularly in systems managing a large fleet or many interdependent processes, this is not always simply a matter of a few clicks.

An important market trend is the ability to manage equipment from different manufacturers through a single higher-level system. The VDA 5050 standard supports this approach by defining the communication interface between transport robots and fleet management systems. It can significantly simplify the integration of devices from different suppliers, but it does not guarantee full plug-and-play compatibility. Configuration, functional checks, testing and validation of the complete system are still required.

The fleet management system coordinates the operation of the mobile robots. It assigns tasks and manages traffic, mission queues, charging and fleet behaviour in exceptional situations. VDA 5050 defines how transport robots communicate with the fleet management system and may therefore make it easier to use equipment from different manufacturers. However, it does not provide full plug-and-play compatibility: every solution still requires configuration, testing and validation in the actual process.

The role of an integrator in a reliable AMR deployment

This is why selecting an experienced integrator is critical. The integrator should do more than supply the equipment: it should design the material flow, docking points, communication architecture, safety measures, mission logic, integration with IT and OT systems, and the subsequent service strategy. At Nextomation, support covers the entire process – from process analysis and concept development through simulation and engineering to commissioning, training and after-sales service.

Which process should be automated first?

Every facility has its own material-flow patterns, spatial constraints and production requirements. The first step should therefore not be the selection of a particular robot, but an analysis of the process and identification of the area in which automation can deliver the greatest value.

During a free consultation, Nextomation experts can provide an initial assessment of the potential for using AMRs, identify a suitable pilot area and determine which data will be required to develop the concept and simulate the system.

Book a free consultation on intralogistics automation

Where can AMR technology be applied?

Some customers still assume that AMR technology is unsuitable for certain types of production or intended only for large, modern factories. In practice, mobile robots are used in sectors including:

• food and beverage,
• electronics,
• automotive,
• pharmaceuticals,
• cosmetics,
• mechanical engineering,
• logistics,
• industrial manufacturing more broadly.

When do AMRs deliver the greatest value?

This does not mean that AMRs are suitable for every process or configuration. They deliver the greatest value when transport tasks are repetitive, there are multiple pick-up and drop-off points, production schedules change frequently, travel distances are significant and operator availability limits the ability to scale production.

The integrator is responsible for configuring the robot, peripheral equipment and control system so that the solution meets the customer’s actual requirements. Simply commissioning the robot and defining a route is not enough. The system must also provide reliable docking, signal exchange, material identification, exception handling, charging management, mission prioritisation and appropriate behaviour in the event of a communication failure.

Safety in a shared environment

Safety is particularly important because an AMR operates in an environment shared with operators, pedestrians, forklift trucks, other transport equipment and machinery. Safety must therefore be verified not only in the documentation, but also through practical testing under real operating conditions.

Robot testing in real-world conditions

A demonstration at the customer’s facility can be used to assess the robot’s ability to localise itself, avoid obstacles, change its route, stop within defined zones and operate in a dynamic environment. Reaction time and stopping distance are not fixed values; they depend on speed, payload, traction, braking-system design and the configuration of the safety fields.

An AMR can detect an obstacle much faster than a person and initiate a safe stop without the delay associated with human reaction time. This does not mean, however, that it will immediately avoid every obstacle and continue moving. The system must first bring the robot to a safe stop. It can then calculate an alternative route or wait until the passage is clear.

Another common test assesses the robot’s ability to leave an area with restricted access or navigate through a dynamically changing arrangement of obstacles. Modern AMRs can recalculate their routes and search for an alternative path, but their effectiveness depends on the physical availability of that path, map quality, localisation accuracy and the configuration of restricted zones.

AMRs versus AGVs – the key differences

AMR technology should not be regarded as entirely new or unproven. Automated guided transport systems have been used in industry for decades, while advances in navigation, sensors, edge data processing and fleet management have significantly expanded the capabilities of modern mobile robots. Deployment has accelerated particularly in response to rising labour costs, the growth of high-mix production and the increasing need to automate internal logistics.

It is important to distinguish between conventional AGV systems and AMRs. Traditionally, AGVs follow predefined routes marked by magnetic strips, markers, inductive wires or fixed navigation points. AMRs, by contrast, use a map of their environment and sensors for localisation and dynamic route planning.

To an outside observer, the two solutions may look similar, but they differ in flexibility. They can be compared to a tram and a car: both provide transport, but a car can drive around an obstacle, whereas a tram remains dependent on its track. The distinction between advanced AGVs and AMRs is becoming less clear, however, because some modern AGVs also use advanced localisation and obstacle-avoidance systems.

Overall, AMRs are particularly valuable in dynamic environments with frequent changes, limited space and restricted areas in which manual transport creates additional organisational and time-related costs.

When to choose a conveyor or a hybrid model

If a production line requires continuous, high-throughput transport of a single component type between two fixed points, a conveyor may be more cost-effective. In such cases, a hybrid model is worth considering: roller or belt conveyors can handle the stable, repetitive part of the flow, while AMRs manage the more variable transport tasks.

Each concept should therefore be designed individually, taking into account actual material flows, transport frequency, production peaks, order structure, available space and the facility’s future development plans.

How is an AMR system implemented in practice?

Stage 1. Client meeting, site audit and material-flow analysis

The first and most important stage is to understand the client’s needs, conduct a site audit and obtain reliable information about material flows. A spaghetti diagram is one of the tools used to visualise the actual routes taken by components, products, containers and employees.

The analysis should cover:

• the number of pick-up and drop-off points,
• the frequency of transport operations,
• the dimensions and weight of the loads,
• waiting times,
• the occurrence of transport peaks,
• the types of containers and load carriers used,
• the width of transport aisles,
• the number of intersections and potential conflict points,
• the condition of the floor,
• changes in floor level and passage through gates,
• requirements relating to lifts, airlocks and controlled areas,
• the methods used to transport empty packaging and waste.

This information can be used to estimate the number of robots required, the necessary charging stations, route lengths, fleet utilisation and a preliminary project budget. Even at this stage, the analysis may reveal hidden problems that are masked in day-to-day operations by manual intervention from operators.

The fleet’s behaviour must also be defined for events such as a fire alarm, evacuation, power outage, network failure, blocked aisle or one robot being taken out of service. These scenarios are often overlooked in preliminary analyses, even though they are critical to safety and operational continuity.

A properly conducted audit reduces uncertainty for both parties. The client gains insight into potential costs and expected outcomes, while the integrator can assess whether the project can be delivered within the planned schedule, budget and proposed technical architecture.

Stage 2. Selection of a pilot area

One of the main advantages of AMR systems is their scalability. Deployment can begin with a single robot serving one clearly defined process. A pilot will not reproduce every operating mode of a large fleet, but it can validate the assumptions, docking method, integration, safety measures and the organisation’s response to the new technology.

The selected process should be relatively simple to launch while still providing meaningful business value. Examples include:

• delivering components from the warehouse to one production line,
• collecting finished products,
• transporting empty containers,
• servicing a single buffer,
• transporting materials between two departments.

Introducing robots in stages gives operators, maintenance personnel, the health and safety team and management time to become familiar with the technology. It also reduces organisational risk and avoids deploying a large fleet into a process that has not yet been sufficiently stabilised.

Stage 3. Simulation and concept validation before committing CAPEX

An experienced integrator should not omit transport simulation, whether the project concerns a pilot area or a full-scale deployment across the entire facility. The client should also expect this analysis because it verifies whether the number of robots, charging stations and buffer points has been selected correctly.

Simulation can be used to assess:

• the utilisation of individual robots,
• the length of mission queues,
• the load on charging stations,
• route availability,
• potential conflict points,
• waiting times at docking stations,
• throughput during peak periods,
• the effect of a single robot failure on the complete system,
• fleet behaviour following a schedule change.

Some areas of a facility may be arranged in such a way that one robot docking temporarily blocks the route for other units. If this is not anticipated, the entire system may be delayed while a single mission is completed. The risk can be reduced by introducing one-way routes, passing points, waiting zones, appropriate task prioritisation and changes to the location of docking points.

A comprehensive simulation should also determine how the system will perform during the morning logistics peak and whether new constraints will emerge anywhere in the flow.

Any savings achieved by skipping simulation are usually illusory. The subsequent cost of correcting routes, redesigning staging areas, adding robots or modifying the control logic may be substantially higher than the cost of modelling the system at an earlier stage.

Modern simulation tools can also account for component flows within production lines. The scope of the analysis depends on the availability of machine models, process data and a digital representation of the facility. The outcome should be more than a visualisation of robot and material movements: it should include a report documenting the assumptions, fleet-utilisation results, risk analysis and design recommendations.

Stage 4. Selecting the solution architecture – work modules, peripheral equipment and integration with higher-level systems

Once the simulation results have been approved and the investment decision has been made, detailed design of the system architecture can begin.

At this stage, the following are defined:

• communication with MES, WMS or ERP,
• the logic used to generate and confirm missions,
• task priorities,
• queue-management rules,
• load-identification methods,
• signal exchange with machines supplied by other vendors,
• the industrial-network architecture,
• cybersecurity requirements,
• data buffering and system behaviour following a loss of connectivity,
• user and access-rights management.

AMR work modules and peripheral equipment are designed in parallel. These may include transport racks, docking stations, conveyors, lifting mechanisms, centring elements and systems that confirm the correct transfer of a load.

The stability of the complete system often depends more on reliable docking and load transfer than on robot navigation itself. An AMR may reach its destination flawlessly, but the mission will still fail if the container is misaligned, the presence of the load carrier is not confirmed or a sensor has shifted.

Stage 5. Safety and operational readiness

At Nextomation, the safety and operational stability of an AMR fleet are treated as fundamental design requirements. Certified safety components on the robot alone are not sufficient. The complete system must be assessed, including the work module, load, routes, docking operations and interactions with people and other vehicles.

At this stage, the required robot behaviour in particularly sensitive areas is defined in collaboration with representatives of the health and safety and maintenance teams. These areas typically include:

• intersections of transport routes,
• pedestrian crossings,
• gates and airlocks,
• areas shared with forklift trucks,
• docking zones,
• narrow corridors,
• areas with restricted visibility.

Depending on the outcome of the risk assessment, the solution may include audible warnings, traffic lights, speed restrictions, one-way traffic, additional sensors, mirrors, floor markings, physical barriers or changes to the routes.

A typical AMR may be equipped with the following safety features:

Safety laser scanners
Depending on the number and location of the devices, safety laser scanners monitor the area in front of, behind or around the robot. They create warning and protective fields:

• when an object enters the warning field, the robot may reduce its speed;
• if the protective field is breached, the robot initiates a safe stop.

The scanning range and the level of safety provided depend on the components used and the system architecture.

Dynamic protective fields
The size and geometry of these fields may change according to speed, direction of travel, steering angle, payload and current operating mode. The higher the speed and the longer the stopping distance, the earlier the system must detect a potential hazard.

Additional obstacle detection sensors
3D cameras, ultrasonic sensors and infrared sensors can support the detection of low, protruding or otherwise hard-to-detect objects that the primary scanner may miss. However, not all such sensors are safety-rated components. Some provide navigation or auxiliary functions, so their role must be clearly defined in the design.

Emergency stop buttons
E-STOP buttons are installed in accessible positions on the robot housing. Pressing a button initiates an emergency stop. Depending on the size of the device, several buttons may be required.

Familiarising employees with AMR technology is an important part of the deployment. Their concerns are understandable, particularly when they are sharing a workspace with an autonomous transport vehicle for the first time.

An analogy with a robotic vacuum cleaner may help explain the basic principle of navigation, but an industrial environment involves very different mass, speed, kinetic energy, loads and safety responsibilities. Industrial AMRs therefore use more advanced detection systems, certified components and clearly defined emergency procedures.

Maintenance personnel also require dedicated training. As a minimum, it should cover:

• basic robot operation,
• safe stopping and restarting,
• alarm diagnostics,
• procedures following a loss of localisation,
• inspection of charging stations,
• responding to blocked routes,
• basic operation of the fleet management system,
• procedures for escalating service requests.

The fleet must also be integrated into the facility’s fire-safety strategy. Depending on the agreed scenario, the robots may stop in a safe location, clear evacuation routes or travel to designated holding areas when an alarm is triggered. There should be no universal default response. The procedure must be based on a risk assessment and agreed with the personnel responsible for fire safety.

Stage 6. Pilot operation, ramp-up and the ‘confidence-building period’

The transition from manual to automated intralogistics does not end with formal acceptance of the system. Experience shows that the first few weeks of fleet operation provide extensive information that cannot be fully anticipated, even through detailed simulation.

The resulting insights are not always technical. New optimisation opportunities often emerge only after the system is running, when users begin to understand its flexibility and identify further processes that could be automated.

Organisational adaptation is crucial during this period. System administrators learn to manage queues, priorities and exceptions. Operators learn how to work alongside the robots, while maintenance personnel gain practical diagnostic experience.

Leaving the client without adequate support immediately after acceptance is a serious mistake. Planning several weeks of post-commissioning assistance makes it possible to:

• adjust route parameters,
• optimise mission priorities,
• eliminate unnecessary stops,
• relocate waiting points,
• refine responses to exceptional situations,
• train users using real operating scenarios.

The project should also clearly define the warranty conditions, service response times, spare-parts availability, remote-support options and post-warranty service arrangements. Not every integrator has its own trained team and the resources required to maintain a fleet over the long term.

The most common mistakes when implementing AMR technology

The most common mistakes across these stages include:

• purchasing equipment without first analysing the material flow and preparing a spaghetti diagram,
• attempting to deploy a complete fleet without a pilot phase,
• failing to simulate and validate the concept before committing to the investment,
• calculating return on investment solely on the basis of headcount reduction,
• failing to integrate the system with the production process and IT and OT systems,
• underestimating the time employees need to adapt,
• failing to establish a plan for training, service and gradual performance ramp-up,
• failing to adapt docking points and infrastructure to actual operating conditions,
• overlooking cybersecurity and access management for the fleet system,
• automating an unstable or poorly organised process.

What benefits can an AMR fleet deliver?

1. Financial benefits

lower internal transport costs,
lower losses caused by downtime and late material deliveries,
lower costs of future modifications and reorganisation than with fixed-infrastructure solutions,
better utilisation of technical and organisational resources,
lower labour costs in demanding areas or zones requiring additional authorisations.

2. Operational benefits

more stable material flow,
more reliable and timely material replenishment,
greater process predictability,
lower risk of bottlenecks,
the ability to change transport priorities dynamically,
improved traceability of completed operations.

3. Organisational benefits

less dependence on the immediate availability of operators,
easier scaling of production volumes,
the ability to introduce automation in stages,
standardised transport operations,
clearer responsibility for individual processes.

4. Safety benefits

fewer manually performed transport movements,
more predictable vehicle behaviour in shared zones,
the ability to record incidents and routes,
lower risk associated with haste, fatigue or restricted visibility.

This does not mean that all risks are eliminated automatically. In a poorly designed system, robots may also cause congestion or create new conflict points. Safety must therefore be assessed across the complete process, not only at the level of an individual device.

5. Strategic benefits

integration of different types of equipment from multiple manufacturers,
easier expansion of the automation ecosystem,
greater resilience to production variability,
preparation of the facility for further digitalisation,
the ability to build a scalable intralogistics architecture.
How quickly can the investment pay for itself?

The payback period for an AMR system depends on the individual application and a range of factors, including:

• the complexity of internal transport operations,
• the number of process points,
• route lengths,
• the types of material transported,
• the number of production shifts,
• current labour costs,
• the frequency and duration of downtime,
• robot utilisation,
• the cost of integration and peripheral equipment.

Many companies currently expect automation projects to achieve payback within approximately two years. This may also be a reasonable target for an AMR project, but it should not be treated as a universal threshold. More complex projects may have a longer payback period, particularly when a significant share of the investment relates to infrastructure modifications, system integration or custom-designed work modules.

A basic payback model should include:

• the current cost of manual transport,
• the cost of shortages and micro-stoppages caused by material unavailability,
• costs associated with safety incidents and risk exposure,
• the cost of work in special or controlled zones,
• the cost of maintaining excessive buffers,
• the cost of future changes to the production layout,
• service, energy, licence and fleet-maintenance costs,
• the value of increased production availability and stability.

The real return on an AMR investment does not come solely from replacing manual labour. In many cases, a larger share of the benefit is generated by fewer disruptions, a more stable production rhythm, more reliable material supply, fewer stoppages and lower costs associated with future organisational changes.

Summary

Internal transport was treated as a supporting process for many years. Today, it is increasingly becoming one of the main constraints on business growth. It can cause delays, increase costs and risk, and make it more difficult to scale production.

Companies are therefore increasingly viewing AMRs not simply as a technology cost, but as a means of stabilising processes, improving safety and creating a predictable flow of materials.

The question should not simply be ‘Should we automate internal transport?’ but rather ‘Which process should we automate first, which risks do we want to reduce and how will we verify the results?’

The best AMR deployments do not begin with the purchase of a fleet. They begin with process analysis and identification of the specific problem the technology is intended to solve. This approach reflects a modern understanding of automation as a tool for managing risk, process stability and production capacity, rather than merely as an investment in equipment.

Strengthening a company’s image as a modern and innovative organisation may be an additional benefit, but it should not be the main justification for the deployment. What matters most are measurable results: more stable production, safer transport, greater flexibility and the ability to scale processes further.

This is why mobile intralogistics is becoming one of the most practical areas of investment in a modern manufacturing plant.