Automated Guided Vehicles and Mobile Machinery

Whether it is a transport vehicle on the shop floor, a mobile robot in manufacturing, or mobile machinery in the field: a machine that moves on its own must perceive its environment reliably. That has been our field of expertise since the company was founded. To achieve this, FusionSystems combines multiple sensors into one picture of the environment and derives from it where the machine is, what is happening around it, and where it can drive safely.

Multi-sensor environment perception as the common foundation

No single sensor sees everything. A camera recognises what an object is, a laser scanner measures precisely how far away it is, and additional sensors provide motion and orientation. Multi-sensor environment perception brings these data together. The result is an environment model that contains people, objects, obstacles, and drivable free space and that holds up even with dust, backlight, or vibrations.

We use the same foundation in road traffic, on the shop floor, and in the field. We describe the methods under Sensor Fusion & Physical AI and the detection of people under Person Detection.

Automated guided vehicles: MULI

MULI is the automated guided vehicle from FusionSystems. We conceived, designed, and developed it ourselves. Its modular structure allows it to be adapted to different tasks; it navigates indoors and outdoors. Software and sensors for person, object, and obstacle detection ensure safe operation in production, logistics, and transport.

An automated guided vehicle makes the material flow flexible: it transports small load carriers between warehouse, machine, and assembly and replaces rigid conveyor technology. We integrate it into the control level through our control and visualisation system.

Mobile robotics: articulated arm on a mobile platform

In the funded project InnoTeam RoSa, we are developing MULI further. The project combines the mobile platform with an articulated robot arm: the robot comes to the workpiece instead of the workpiece being brought to the robot. For this, the system must capture its environment in three dimensions, know its own position precisely, and coordinate the motion of platform and arm. Our environment perception provides the foundation for this.

More about the project is available under Funded Projects.

Mobile machinery: environment perception in the field

What applies on the shop floor also applies outdoors. In the KoSiNuS project, we are equipping an agricultural machine with vehicle environment perception. We process and fuse the data from various sensors so that the machine reliably detects people, obstacles, and a clear travel path, between rows of plants as well as at the edge of the field. The project develops central components and safety architectures for autonomous mobile machinery.

Learn more in the FusionBLOG and under Funded Projects.

Navigation and localisation

A mobile machine must know where it is at all times. We use several types of navigation for this and combine them into hybrid navigation:

• Line guidance: The machine follows a line. This is easy to set up and insensitive to uneven floors and vibrations.
• Landmarks: A camera determines position and orientation using markers on the ceiling or wall. The routes remain freely selectable and can be changed at any time.
• Contours: A laser scanner matches the outlines of the environment against a map. The method is also suitable for outdoor areas.

The combination increases efficiency, flexibility, and safety. We set up the sensors for your environment.

Participation in technical committees

FusionSystems contributes its experience to the professional community. Dr.-Ing. Ullrich Scheunert and Axel Pätzold are active members of the VDI technical committee “Autonomous Mobile Robotics”. FusionSystems is a member of the industry association Forum-FTS.

We describe the application in manufacturing under Manufacturing and Mechanical Engineering and environment perception in road traffic under Environment Perception for Autonomous Public Transport.