CONNECT WITH US

0% under reject - Keeping an AOI on autonomous driving's future

Press release 0

Artificial Intelligence (AI) and Advanced Assisted Driving System (ADAS) are the two core elements of the Automated Vehicle (AV). A whole battery of sensors comprises the ADAS. These include short and long-range radar, LiDAR & precision cameras, which help to collect information outside the car at real time. With the help of the AI, the ADAS is then able to differentiate sensor data into crucial categories including pedestrians, obstacles, traffic signs, lane line, other vehicles, proximity, speed and direction relative to the AV, process this and quickly instruct the vehicle's actuators that control acceleration, braking, and steering.

"Since sensors play such a crucial role in enabling AVs, a critical component of delivering the most reliable AVs is to make sure that every sensor produced during manufacturing is free from defects," said Dr. James Song, Vice President and Product General Manager of ASM Pacific Technology (ASMPT). Optical-based sensors such as LiDAR and cameras are no exception. A tiny particle on a LiDAR or camera sensor surface could cause huge issues due to the signal not able to be received or transmitted, which could have huge implications for AV reliability and passenger or pedestrian safety.

Dr. James Song emphasized safety is the primary concern in the automotive industry. With more complex camera assemblies, the need to ensure the best quality from the manufacturing process becomes the most critical. It has indicated that there is an increasing market demand for its Automated Optical Inspection (AOI) machines to replace conventional inspection methods that employ human operators. With its automated visual inspection of component manufacture using a camera that autonomously scans the device under test for both catastrophic failure and quality defects, AOI is overall much more reliable and repeatable than manual visual inspection.

Recently, ASM Pacific Technology has lunched its brand new AOI model named "CamSpector," which has been designed specifically to meet automotive market needs, and to fulfill important requirements such as CE Mark. It is the first 3rd Optical AOI capable of detecting sub-micron particle as well as 2D + 3D wire bond defects for the CIS market. CamSpector equipped with HD optics, which enables a more robust vision performance to reduce over reject rates to less than 0.1%, and to keep under reject at 0%. The "secret sauce" behind these impressive capabilities are Artificial Intelligence capabilities that have been integrated into CamSpector's AOI vision system to leverage big data from both inspection and manufacturing & enterprise systems.

With manufacturers increasingly seeking to modernize their factory lines with more automation and AI features in tandem with "Industry 4.0" trends, AOI inspection data can be shared to such "Smart Factory" systems to augment their data analysis. With real-time feedback systems a Smart Factory hallmark, AOI data could help to improve yield and reduce operational costs, or on a wider level even be used to monitor and provide feedback for upstream processes in the Smart Factory so as to proactive repair of faulty processes or improvement process capability.

High quality camera senor is critical to the reliability of AV. Therefore, a production line of the sensor equipped with intelligent and high-precision inspection system is prerequisite for achieving the autonomous driving. "This is where CamSpector comes to play," said Dr. James Song. It is expected that more manufacturers will integrate AOI system in the production line of auto parts in order to improve product quality. These are among the various possibilities as the world becomes more enabled via an increasingly integrated approach blending intelligent software with precision manufacturing systems.

Multiple sensors

Multiple sensors
Source: Yole 3D Imaging & Sensing 2020 Report

ASMPT's AIoT

ASMPT's AIoT