We develop optimized customer-specific solutions

Realistic simulation of the enemy during military exercises.
The Polytronic Robotic Target enables end users to conduct realistic, scenario-based live-fire training activities in various environments.

Stettbacher Signal Processing AG (SSP) not only has the necessary expertise for the industrial use of artificial intelligence (AI), but also extensive experience and the necessary cameras (b/w and color, mono and stereo) and lenses, as well as proprietary programs and software for AI and classic image processing. The latter can be adapted and trained very quickly and easily to any specific requirements.

The image shows the online detection and classification of mechanical and electrical elements in real time. Note how early the objects are detected, even when they are only partially visible. The objects do not have to be presented in a specific position, but can lie or stand anywhere, even partially covered. The system can also determine and communicate the location and position of each object, for example to implement pick-and-place applications, such as those using industrial robots. Please contact us if you have any questions.

Following recent talk of an “AI race” between Huawei and Nvidia, we would like to add that we at Stettbacher Signal Processing AG (SSP) use Nvidia technology. The image shows the series production of our stereo camera, which integrates an Nvidia Jetson Xavier NX module with 384 CUDA cores, 48 tensor cores, and 16 GB RAM. The module, which is barely larger than a credit card, is located on the back of the camera and contacts the heat sink integrated into the housing shell.

This enables the stereo camera to recognize objects in real time and precisely locate and/or measure them in its 3D coordinate system. Thanks to the Linux (Ubuntu) system running on the Nvidia module, users can develop their own applications for the camera. The camera’s solid, waterproof housing allows it to be used outdoors, in harsh environments, and even underwater. We would be happy to support or advise you on your solution.

The controller for autonomous vehicles from Stettbacher Signal Processing AG (SSP) offers everything needed for autonomous driving – whether on land, water, or in the air, as well as indoors and outdoors. Solutions are already available for all essential basic functions:
(*) Pose Estimation: Determining position, location, and orientation based on GPS, UWB, compass, inertial, etc.
(*) Map Handling & Path Planning: Planning trips to a destination, either across terrain with or without obstacles or on permitted paths, roads, etc.
(*) Path Driving: Driving along a planned path (path control) while stopping in front of obstacles, driving around them, or planning a detour on the fly (collision avoidance).
(*) Communication: The controller also has various interfaces (CAN, RS486, Ethernet, etc.) through which it can communicate with the user’s subsystems. It also has network/WiFi, 4G/LTE, etc., so that it can remain connected to the user or a control center at all times, even over long distances, indoors and outdoors.

The corresponding sensors, modules, and antennas are integrated into the controller or can be connected directly. The controller allows autonomous solutions to be implemented quickly and elegantly. In addition, with SSP, you have an experienced partner at your side who you can call on at any time if necessary. A classic win-win situation!

Our customer was in a dilemma: his partner for the automation of his device repeatedly failed to control a drive in such a way that the required dynamics and precision of a machine movement could be achieved. Our analysis (Stettbacher Signal Processing AG, SSP) quickly showed that neither the PLC nor the control algorithm used were capable of meeting the requirements.

We then developed an optimized solution on behalf of the customer. To do this, we designed and simulated a model-based control algorithm that easily ensured the required movement. We then implemented the hardware solution shown in the figure with integrated power supply, communication, etc. (see figure), based on a microcontroller that, even when using the new algorithm, can control more than ten times faster than the original PLC and is also significantly more cost-effective in series production. The customer was then ready to launch its machine on the market.