News Release

Control of MEMS lidar mirrors enables steady pictures in shaky environments

Control of the MEMS mirror paves the way for automotive applications

Peer-Reviewed Publication

SPIE--International Society for Optics and Photonics

Experimental investigation of vibration suppression for MEMS lidars.

image: The observed 3D scene (top left), the measurement results without external vibrations (left), the results with vibrations but no controller (middle) and the results with vibration and a MEMS mirror controller (right). The distortions caused by the vibrations are clearly visible in four distinct features of the scene (marked by white squares). view more 

Credit: Woo et al. doi: 10.1117/1.JOM.2.1.011007.

3D object detection in various road situations is a necessity for safe autonomous driving. For example, a white 18-wheel-truck next to the cloudy sky or a single, dropped, black tire on a highway needs to be detected early enough to avoid collision. Lidar (light detection and ranging) is one of the key sensor technologies that will enable a reliable perception of a car’s surroundings due to the high resolution and long viewing distance that can be achieved.

However, lidar is not yet widely adopted in commercial automobiles, which is mainly due to the high unit price and its limited lifetime due to the mechanically rotating parts that are needed for scanning the road ahead. To make lidar sensors a standard safety feature, similar to airbags, MEMS (micro-electromechanical systems) scanning mirrors are regarded as the game changer that will enable low cost and long-term stable lidars. A tiny mirror, with a couple of millimeters in diameter, is patterned on a silicon body and made to oscillate thousands of times per second to steer the laser beam. This replaces the bulky and expensive rotating motor and bearings. However, the influence of vibrations, which are caused by bumpy roads and driving in general, on the performance of MEMS lidar measurements have not been reported so far.

To close this knowledge gap, a team led by Prof. Georg Schitter of Technical University of Vienna (TU Wien) and Leonhard Kormann from Infineon Technologies Austria AG, supported by the Austrian Research Promotion Agency (FFG), aim to directly evaluate the robustness of a long-range MEMS scanning lidar against external vibrations. In their research reported in the Journal of Optical Microsystems, they use a long-range MEMS lidar prototype, which is able to detect a tire on the road from 130 m, that is equipped with an application specified integrated circuit (ASIC), which senses the timing of the MEMS mirror movements with nanoseconds accuracy and secures a stable oscillation by a digital controller.

Experimental results reveal that the MEMS control in the ASIC can successfully suppress the errors caused by vibrations. For the MEMS lidar without the controller, the applied random vibration distorts the 3D measurements along the scan axis, leading to errors in the 3D surfaces and small, flying particle-like points around the edges of the objects. When the control of the MEMS mirror is turned on, these errors are reduced significantly, allowing for reliable 3D object detection and ranging. “The MEMS controller reduces the influences of innate vibrations with automotive lidars and maintains stable 3D images, enabling reliable imaging processing and object detection,” Han Woong Yoo, a lead researcher for MEMS control and lidars, said about the achieved performance. “A robust control design for the MEMS mirror is the key technique for operation under vibrations and shock, which will provide affordable and highly reliable MEMS lidars in automotive use.”

In summary, MEMS lidars have large potential in automotive application and could bring enhanced safety in future mobility with an attractive price.

Read the open access report by Yoo et al.: "Evaluation of Robustness against External Vibrations for Long-range 1D scanning MEMS Lidar," J. Opt. Microsyst. 2(1) 011007 (2022). doi: 10.1117/1.JOM.2.1.011007.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.