The announcement of these three inventions has sparked significant interest and garnered favorable responses from esteemed academic and industrial experts.
Professor Jaka Sodnik, from the University of Ljubljana in Slovenia, commended Dr. Duong’s research group for their significant contribution in tackling the autonomous driving challenge in urban environments. Their work’s comprehensiveness and practicality in a large-scale context mark a noteworthy achievement.
According to Japhael Trindade, the chief architect at Huawei in Germany, this solution offers an affordable option that can be implemented in both high-end and low-end cars. This solution addresses the current limitations that are hindering the widespread adoption of autonomous cars or systems.
Exploring the Current Issues, Challenges, and Solutions of Self-Driving Cars
The concept of the “intelligent car” is not new and has garnered significant attention from the research community and substantial investment from industry over the past two decades.
In 2018, a breakthrough in automotive technology occurred with the introduction of a semi-automatic vehicle that possessed lane-keeping and lane-changing features. Notably, Mercedes-Benz made further advancements in 2021 with the launch of their S-Class model. This innovative vehicle has the capability to drive entirely autonomously, reaching speeds of up to 60km per hour, particularly in congested traffic situations. This groundbreaking achievement establishes the S-Class as the world’s first level-3 fully automated vehicle, representing a significant milestone in the evolution of automotive technology.
According to the latest automotive report from Forbes in February, test vehicles from companies such as Waymo and GM Cruise have achieved a significant milestone of over one million miles in fully autonomous urban driving, without the need for human supervision. Additionally, these vehicles have covered a remarkable six million miles with the assistance of human safety monitors.
Although these solutions have the potential for success, their practicality is not guaranteed due to the significant investment and production costs involved.
The cost of self-driving technology is prohibitively expensive for individual car owners, and it also poses significant financial challenges for companies offering autonomous vehicle services. For instance, Waymo incurred a loss exceeding $5 billion in 2021, while GM Cruise had to spend up to $5 million per day to implement its self-driving system in the densely populated city of San Francisco.
At present, fully autonomous vehicles are not yet widely available on roadways and significant progress is still required in the realm of self-driving cars.
Autonomous driving stands as one of the most complex challenges that mankind has faced. It demands the integration of advanced technologies from numerous industries including robotics, artificial intelligence (AI), automation, materials, mechanics, computer science, and more.
The self-driving car conundrum presents a significantly greater challenge in urban settings compared to highways due to the multitude of diverse obstacles and complex scenarios that need to be managed. As a result, the automotive industry is faced with a formidable task.
Dr. Duong aims to develop a scalable solution with a camera-based core technology, instead of following the high-cost approach taken by Waymo and GM Cruise in order to solve the problem.
In his presentation, he showcased three significant inventions in the field. These inventions are as follows:
1. Three-dimensional sensing: This technology allows for advanced sensing capabilities in three dimensions.
2. Automatic calibration: This invention automates the calibration process, enhancing efficiency and accuracy.
3. Asynchronous map updates for multiple sensors: This integrated solution enables the creation of detailed maps by seamlessly combining data from multiple sensors with asynchronous updates.
These inventions offer groundbreaking solutions that can revolutionize various industries.
Leveraging both inventions (1) and (2) alongside AI-driven object recognition solutions, our esteemed research team has successfully developed a sophisticated level 2+ (semi-automated) vehicle system architecture, making use of camera technology.
The product innovation has the capability to enhance the system by enabling it to reach level 3 and level 4 when integrated with more advanced sensors.
Our solution is designed to cater to a wide range of vehicle models, including both affordable options and luxury vehicles equipped with high-end sensors such as Radar and Lidar.
|Dr. Nguyen Van Duong is the opening speaker at the Innovation Connectivity Autonomuos – ICA Summit 2022 held on May 30-31, 2022 in Munich, Germany. Photo: Supplied|
In recent times, leading automotive manufacturers including Daimler AG (Germany), BMW (Germany), Tesla (USA), and NIO (China) have implemented AI recognition technology with remarkable outcomes.
According to Dr. Duong, NIO has the potential to become a major industry player if they maintain consistent technological development and offer competitive prices on a larger scale in the long term.
According to Dr. Duong, NIO’s vehicles are characterized by their superior quality and advanced functionalities, surpassing other models in the same market segment. This has led to heightened competition among rival European and American automakers.
The Mercedes-Benz S-Class has a higher price point due to the inclusion of sensors, ECUs, and luxury components, which make it exceed the US$200,000 mark. In contrast, NIO’s offering is available at a more affordable price, approximately half of the S-Class’ cost.
According to Dr. Duong, NIO’s autonomous vehicles may feature high-end sensors that are on par with or potentially even more costly than those found in the Mercedes-Benz S-Class model.
From a strategic standpoint, it is very plausible that NIO’s financial losses are intentionally incurred to significantly enhance customer appeal.
Tesla, the innovative automotive company, has set its sights on developing cost-effective solutions for automated vehicles. In 2022, Tesla made a strategic decision to prioritize the advancement of its camera system technology. This move aims to make autonomous driving more accessible and affordable for consumers. Tesla’s dedication to innovation and commitment to improving the future of transportation is evident in their proactive approach to automated vehicle development.
While AI has made significant advancements in object detection and environmental sensing, solely relying on it may not be a sufficient solution. Urban traffic presents numerous challenges where existing AI solutions have limitations in recognizing certain objects.
According to industry experts, Tesla’s AI solution has raised concerns due to a high number of accidents in the past two years. In total, there have been 736 reported accidents, indicating potential doubts about the effectiveness of Tesla’s autonomous driving technology.
According to Dr. Duong, Tesla’s approach in the race for safe and sustainable automated technologies cannot be deemed rational.
The Vietnamese scientist stated that, apart from leveraging artificial intelligence (AI), his team has developed three groundbreaking inventions aimed at enhancing the comprehension of car systems regarding their immediate surroundings.
“Our system architecture allows for semi-automatic (level 2+) operation using just the camera, while level 3 and level 4 require additional sensors for enhanced safety,” explained Dr. Duong. “Therefore, I believe that our flexible and extensible approach will enable us to achieve even greater benefits on a larger scale.”
Dr. Duong, highly respected as a leading scientist and automotive industry expert in Europe, expressed his anticipation for fostering collaboration between Vietnamese and European counterparts. With a track record of numerous significant contributions, Dr. Duong aims to leverage his expertise towards the advancement of both parties.
“While evaluating projects for funding by the EU Commission in April, I had the opportunity to collaborate with several universities, research institutes, and companies in Vietnam,” Dr. Duong shared. “We identified promising candidates who successfully advanced to the final round. Regrettably, the ultimate decision was entrusted to a different country.”
However, Dr. Duong regarded this experience as an initial effort to enhance the visibility of Vietnamese candidates, thereby creating greater prospects in the foreseeable future.
Dr. Nguyen Van Duong
Head of ADAS Department
Panasonic Systems Automotive Europe (PASEU)
Representative at EU Commission
The individual has been elected as a senior member of the IEEE Society, along with being appointed as a member of various prestigious IEEE groups, such as IEEE P2020 Automotive, IEEE Robotics, IEEE ITS, and EU ITS.
In addition to his roles as a researcher and professor, he also holds positions as an editor and advisory board member for various prestigious journals and conferences.
- CARAMEL (2019)
- CPSOSAWARE (2020)
- TRUSTEE (2022)
- BERTHA (2023)
- ZEV-UP (2023)