Ian Tzeng, product strategy and marketing manager at oToBrite Electronics. Credit: DIGITIMES
Demand for vehicle-to-everything (V2X) and advanced driver assistance systems (ADAS) is growing in the car-use electronics sector. Taiwan-based manufacturers are looking to seize these opportunities.
Over 50% of exports from Taiwan-based manufacturers in the car-use electronics sector are of driver information systems (DIS), followed by 20% for engine transmission systems, according to data from Industrial Technology Research Institute (ITRI) and Automotive Research and Testing Center (ARTC). DISs have benefited from smart car business opportunities.
Demand for head-up displays (HUD), panels and audio-visual systems is also expected to significantly increase. Research firm Strategy Analytics estimates the total production value of automotive electronics will reach US$355 billion in 2030, compared to US$267.6 billion in 2018, at a compound annual growth rate (CAGR) of 5.6%.
Ian Tzeng, product strategy and marketing manager at oToBrite Electronics, pointed out that electrification and intelligence are trends all major automakers and supply chains must follow. In the past, tier-1 suppliers in the carmaking industry were familiar with the manufacturing but not software-hardware integration, which can now be used as an entry point for ADAS vendors.
Currently, there are three strategies being used by passenger vehicle makers to develop self-driving systems. The first is to do the R&D themselves. Automakers such as Tesla and XPeng Motors that have the funds and software-minded talent can accelerate the implementation of self-driving technology through agile development. The second is to partner with a tier-1 manufacturers; most traditional automakers choose to adopt the same supply chain management model as before. The last is to work with highly-customized tier-1 companies or alliances, such as oToBrite and MIH.
Tzeng said the traditional passenger vehicle supply chain is inherently complex and the supply chain management is very cautious. Typically, a single automaker has a limited number of suppliers that they work with for 10-20 years. This structure makes it difficult for manufacturers outside of the supply chain to enter, even if they have advanced technologies and stable output. Therefore, ADAS startups that want to get into the supply chain have a better chance if they work with tier-1 manufacturers.
Tier-1 companies have an advantage when it comes to design and manufacturing. However, they lack the core capabilities in software and AI algorithm. As such, the transformation process has been difficult on the auto industry. It is why tier-1 companies are willing to work with companies that have these solutions. This is also the reason oToBrite uses its vision AI intellectual property (IP) rights to approach tier-1 companies.
Tzeng said breaking into the tier-1 ecosystem is difficult. To do so, companies must first put aside their competition. In the past, tier-1 companies had the advantage of economies of scale in hardware manufacturing. Prices were generally low which made it difficult for those outside the supply chain to compete. As a result, unless tier-1 companies are aware that their own software development capabilities cannot meet the expectations of the market or automakers, breaking into tier-1 supply chains will still be difficult.
It is also very important for tier-1 suppliers of international automakers to have their products obtain international certifications. Automakers pay particular attention to safety and the process. If a product cannot provide IATF 16949 international standard documentation, it will be difficult to find cooperation opportunities in the future.
Tzeng said when discussing cooperation with tier-1 companies, they only care about whether or not you can do a real car demo now. This is a high threshold for many companies that have the technology but have yet to actually put it into practice.
Additionally, since tier-1 companies service global automakers, they pay close attention to the stability, safety and continuity of the product supply. Many will require partners to provide algorithm source codes to ensure when future models need to be optimized and trained there is no risk of the partner skipping out. This way it can continue to stably provide automakers with good quality and guaranteed services. However, for the partner, this type of agreement has risks.
Based on this, oToBrite currently provides different solutions to customers with different requirements. For example, it authorizes vision AI IP for tier-1 manufacturers to fill the gaps in their software development. For startup automakers, it provides electronic control units (ECU), photographic lenses and AI algorithms for ADAS. For IPC factories and limited-area commercial vehicle markets, it mainly focuses on high-end photographic lenses.
In terms of overall revenue, ADAS still accounts for 90% of oToBrite’s total revenue. Vision AI IP and photographic lenses account for less than 10%. Tzeng pointed out that these three product lines will continue to expand in 2022, which is expected to balance out the current one-sided revenue share.
