Development of an Artificial Weather Chamber that Reproduces a Dynamic Weather Environment for Autonomous Driving Sensors (2nd Report)

Haruki Seto･Hiroyuki Enoki･Hirokazu Tanaka (Esoec)

For the purpose of evaluating ADAS functions and performance, we have developed a new all-weather artificial weather chamber that can reproduce rainfall, fog, and snowfall.
We have used this chamber to evaluate visibility in rain, fog, and snow conditions.
In this report, we collected field data such as rainfall, fog, and snowfall, and examined technological development and evaluation methods for reproducing it.

Proposal of Free Space Evaluation Method Considering Sensor Error Characteristics and Consistency of Detection Criteria

Toshiyuki Adachi･Hitoshi Hayakawa･Yuji Oishi (Hitachi)･Yoshinobu Ogasawara･Shigenori Hayase (Hitachi Astemo)

In free-space detection for the realization of high-performance ADAS, there was a problem that the quantitative evaluation differed from the qualitative evaluation based on visual observation. In this manuscript, we present a correspondence-based evaluation method that considers sensor error characteristics in both the circumferential and radial directions to reduce the discrepancy in evaluation results.

Smart Battery Health Algorithm
-Advanced Diagnosis of 12V Lead-Acid Vehicle Batteries using a Machine Learning Approach-

Takehiro Ogawa･Ramirez Bernard･Duhart Bronson･Diaz Moises･Molinar José (Continental Automotive)

Continental’s AI based algorithm predicts battery failures and alerts users.
Collect Data by using battery sensor installed in vehicles and transfer data to cloud via OEM telematics unit or OBD dongle.
An enhanced algorithm based on cloud data for better detection & prediction of battery failures.