Advanced Driver Assistance Systems (ADAS) are expected to become standard features in ten years, and the demand for semiconductors in the systems will be growing exponentially, experts said.
According to data from Goldman Sachs, 90% of new cars in the next decade will be equipped with ADAS. That means ADAS features might become standard in various vehicles before EVs become widely popular and that there will be rapid growth in the demand for chips and sensors driving the systems.
Along with this trend, the value of semiconductors used in ADAS solutions will be rising as well, industry experts said.
Taiwan Taxi, the leading taxi service supplier in Taiwan, has said it is working with AI tech supplier OmniEyes to develop an ADAS solution, hoping to create more smart services.
They are building an AI model to analyze the images collected by taxis' driving video recorders, training the model to determine whether the driver is breaking any rules and to recommend the best possible routes for the driver to avoid wasting fuel.
Taiwan Taxi is taking preemptive action as it anticipates the growing importance of ADASs, industry observes said, adding that vehicle-to-everything (V2X) is a critical part of smart mobility.
About 50 of Taiwan Taxi's 22,000 cars have been equipped with ADASs, while the company looks to further cooperation with OmniEyes this year, the taxi fleet said. The two sides have not yet reached a conclusion regarding the exact scope of their partnership. Meanwhile, Taiwan Taxi is working with a startup offering graphical information regarding traffic images, it said.
ADAS development is necessary for building up smart transport systems. Through sensors installed in and around a car, an ADAS can monitor tire pressure, signal about reverse motion and motion that strays from appropriate lanes, and collect all data for further analysis.
Industry observers noted that ADAS developers in Taiwan are mainly working on panorama monitoring, image sensors, or automotive chips. While ADASs have been identified as indispensable features for future cars, developers have to tackle three challenges – how to manage the cost of lower-end ADAS installed on cheaper cars, how to apply autonomous driving data to an ADAS, and how to manage ADASs on software-defined vehicles (SDVs), suppliers said.
Slow vehicles are a good starting point, said Turing Drive business development director Stephen Liu.
There is a wide range of traffic scenarios in Taiwan, so the nation can provide a pilot testing ground for other regions that are also densely populated, Liu said.
Turing Drive's primary goal in 2022 is to develop smart driving services on slow vehicles like golf buggies, while working gradually toward autonomous driving, he said, expecting driverless systems with SAE Level 2 and 3 to become standard features.
The company is integrating resources in its vertical sectors, such as selling smart driving systems to slow vehicle OEMs and offering slow vehicle driving services to amusement park operators, Liu said.