Mindtronic uses machine learning technology to create a personalized driving experience

TAIPEI, Nov. 8, 2022 /PRNewswire/-- This article is based on an interview undertaken by FusionMedium's  technology online media, TechOrange, and published with permission:

Make cars more than just a "phone with four wheels"!

Make cars more than just a "phone with four wheels"! Mike Huang, CTO of Mindtronic AI, creates a personalized driving experience with machine learning technology. (PRNewsfoto/Fusionmedium)

According to data from the Taiwan Institute of Economic Research, driven by market demand and by government policies put in place at all levels worldwide for the electric vehicle (EV) sector, global shipments of EVs in 2021 exceeded 5 million units, while the penetration rate is estimated to reach 20% by 2025. According to Boston Consulting Group, the value of the global autonomous vehicle market is expected to increase to US$42 billion by 2025, with Level 1 and Level 2 vehicles that incorporate self-driving and intelligent technologies on track to account for about 10% of the overall vehicle market.

The market scale of smart vehicles continues to multiply, making electric and self-driving vehicles an irreversible trend. An intelligent vehicle system includes, in addition to ADAS and self-driving systems that assist driving judgment and enhance safety, an intelligent cockpit composed of a smart console, instrument panels, and on-board entertainment systems. With Taiwan at the forefront of the emergence and development of the ICT and semiconductor industries over the past few decades, the island's vendors are well-positioned to break through the existing barriers and create greater market opportunities as multinationals reconstruct the supply chain for EVs.

The transformation of the automobile has led to the heightening of the importance of software-defined hardware. For example, the ADAS, self-driving, and intelligent cockpit systems that are being developed by major automakers all show that drivers value software functions and user experience over hardware. Under this premise, apart from the dominant position of Taiwan's semiconductor and ICT industries, what advantages do the island's start-ups have in terms of software and hardware integration and software development?

Does Taiwan have the advantage in creating self-driving cars? Three conditions are indispensable!

In the past few years, Taiwan has been aggressively consolidating semiconductor design, manufacturing, packaging, testing, software, and ICT system companies to vertically integrate upstream and downstream industry chains, as well as connect to international cloud leaders, electronic design automation software leaders, and potential start-ups, to form cross-domain alliances that power Taiwan's auto industry with advantages in the next wave of the smart revolution.

Looking at the progress being made by Taiwan's automakers vis-à-vis its foreign rivals when it comes to advanced driver assistance systems (ADAS) and self-driving systems, Mindtronic AI Co. Ltd, which has already garnered the CES New Innovation Award three times for its intelligent cockpit solutions, is a start-up company in the island's smart car sector with growth potential in the international arena.

When analyzing the development prospects and trends in the global self-driving technology market, Mike Huang, Mindtronic's Chief Technology Officer, is of the opinion that there are three key conditions that need to be fully met in order for Taiwan to become a self-driving technology powerhouse. They are marketing, funding, and the possibility of commercialization. To create an environment suitable for the development of self-driving cars, it is necessary to do three things. Firstly, as the R&D of self-driving cars often need tens of billions of dollars in investment, it is necessary to have a sufficient number of investors and the associated funds in place. Secondly, since countries have their different development policies and norms for self-driving cars in various stages from R&D to launch, it is not feasible to "develop a model in place A and test it in place B". It is necessary to have sufficient support from the domestic market. Thirdly, in order to truly achieve a software-defined vehicle (SDV), it is necessary to have sufficient market support. Finally, to truly enable SDV, one needs a complete software development environment.

Huang believes that Taiwan's advantage in the era of self-driving cars is talent. Despite the challenges, Taiwan's small but solid industrial environment is well-positioned to deliver highly concentrated resources for IP start-ups. "Imagine that Taiwan is such a small island with so much talent spread across the electrical, information, mechanical, aerospace, and engineering sectors. With the majority of graduates gathering in this small place, the clustering effect is quite significant." Having such a concentration of the right talent is a unique condition for the development of start-ups.

What value lies in store for the car of tomorrow? Mike Huang: A personalized driving experience is the key!

What is the most critical element for developing self-driving technologies? The most important question to ask when thinking about the nature of smart cars and self-driving cars is "Why does a person need to buy a self-driving car?" and "What do consumers really want?"

The popularity of cell phones has brought people new ways of accessing information, better user interfaces, and more humanized human-machine interfaces and experiences. If we want to provide the same consumer experience in the mobility service industry, it is more important to think about how to deliver an experience beyond cell phones to users. "What is needed here is not one modular function after another, but how to provide a central idea with the ultimate goal of 'personalization'."

For example, the 3D AR HUD (Head-Up Display with Stereoscopic Augmented Reality) technology developed by Mindtronic can project navigation routes directly onto the road ahead rather than on the windshield. The AR technology superimposes virtual objects on the real road to create an integrated virtual-reality experience and an innovative driver experience.

What is a 'personalized' user experience? Huang points out that it is not difficult to repurpose the voice assistant into the car, but it is not as simple as setting up a password or asking the assistant to perform an action. The role of AI in a truly personalized experience is to filter data and categorize messages that users may not have even discovered. AI learns the driver's user behaviors and makes choices for him or her to truly personalize the cockpit and the experience delivered by that cockpit.

A good user experience comes from the AI setting up exactly what the driver wants. In the cockpit environment, it may be about the state of the driver, the positioning of the seat, the weather, the view field on the road, the look on the driver's face when reacting, and the focus of the driver's eyes. With such multiple and complex relationships, there is no way to collect such a huge trove of user information at once and create a personalized model that is applicable to everyone.

AI creates an intelligent driving experience, and knowledge intensive products show the value of software

The value of Mindtronic. "The most difficult part of human-robot collaboration (HRC) is that every user is different. To solve this dilemma, it is necessary to deploy self-learning technology, to enable the machine at the Edge to learn by itself as it works with the user." Creating a driving experience that best suits the driver can also be done by detecting the driver's eyes and, with the accumulation of data over time, learning what the driver is interested in. Mindtronic is now developing a tour-guide robot that will automatically describe the features of a place of interest when the car's AI determines that the driver is particularly interested in it, transforming the meaning and the experience of the classic road trip.

Such data and information have great commercial value. "When a consumer is not interested in a product but, nevertheless, receives a DM, it is advertising. In contrast, when a consumer is interested in a product and, as a result of that interest having been identified, immediately receives a discount coupon, it is a service. Huang believes that in the future, the smart cockpit of self-driving cars will be able to interact with cell phones or personal accounts to create a more complete user experience.

"The value of the car of tomorrow lies not in how much horsepower the vehicle has or the specifications of the hardware components, but in whether the user feels a new and intelligent experience the moment he or she gets seated behind the wheel," Huang emphasized that intelligent cockpits are still in the integration stage. In the past, the instrument panel, central console, and head-up display were all separate and independent. In the future, when all in-cabin features including seats, audio, navigation, streaming, ADAS, gaming, and Internet access are controlled by a single piece of hardware, the value of AI will emerge.

Mindtronic looks forward to forming an industrial ecosystem with more name brands, to leverage the innovative capacities of small companies and the abundant resources of large ones to complement each other. For Taiwanese start-ups, strategic alliances are the best way to develop the automotive industry chain. In addition, Mindtronic is also actively connecting to Taiwan's AI ecosystem horizontally by joining the AI on Chip Industrial Cooperation Strategic Alliance established by the Smart Electronics Industry Project Promotion Office (SIPO) of the Industrial Development Bureau, Ministry of Economic Affairs. Through the alliance, Mindtronic will help connect upstream and downstream partners on the industry chain to create business opportunities and encourage more of the island's manufacturers to invest in the AI on Chip sector, to enable the whole of Taiwan's AI industry to demonstrate its competitiveness in the global market.

SOURCE Fusionmedium