Top list of IFOY nominees (1): Current approaches to vehicle localization in intralogistics are based on UWB (ultra-wideband) or LIDAR (light detection and ranging). Both are relatively expensive. Synaos IMS utilizes additional cameras and a technology called "Visual Odometry".
The technology has been known for some time in the fields of robotics and autonomous driving, but according to the manufacturer, this is its first use in logistics. Synaos also uses simple markers at frequently used locations within the warehouse to provide the system with a reference point for distance measurement. No custom map needs to be created to achieve good results. Only a few markers are required, which can be placed approximately every 100 meters to improve accuracy. According to Synaos, the environment could change at any time without affecting the localization.
The system utilizes advanced vision technology – a camera-based sensor kit for the cost-effective localization of manually driven vehicles. Even with manually controlled transport units, numerous new opportunities arise once their position is known. Key performance indicators such as fleet utilization can be recorded, heat maps created, and layouts optimized. Furthermore, the system enables optimal navigation for drivers through factory halls and logistics centers, allowing for the early detection of delays. In addition, flexible, near-real-time transport order planning can be implemented. Scan-free processes are also made possible, further simplifying operations.
So far, these and other use cases are rarely, if ever, implemented – existing localization solutions are often simply too expensive. As a result, the potential savings are limited compared to the necessary effort. Synaos therefore pursues a modern approach: intelligent software combined with cost-effective hardware. The installation kit houses a stereo camera and a powerful embedded computer. This comparatively inexpensive combination enables the implementation of use cases that were previously unprofitable.
The sensor kit allows entire fleets to be retrofitted with digital tracking. The market for this solution is lucrative: approximately 1.5 million forklifts were produced in 2020. This equates to a correspondingly large number of existing vehicles. However, the location of only a small fraction of these vehicles is currently tracked, nor are they digitally mapped.
The Fraunhofer engineers assessed the upgrading of manual transport systems with the localization solution as highly relevant for mixed operation of autonomous and manual vehicles during the pre-check. Despite a comparatively low price, the solution achieves very high accuracy. The vehicle's position is calculated from the camera images using robust algorithms. Apart from an accelerometer, no other input sources are required. All image processing takes place locally on the embedded computer – no sensitive data needs to be transferred to any cloud. The image data captured by the camera is discarded immediately after processing.
https://www.ifoy.org/bewerbungen/2013/1142/VEHICLE%20LOCALIZATION%20Data%20Sheet%20EN%20v2.pdf
