September 29, 2021, Wuhan – Wuhan Economic Development Zone and Dongfeng company jointly held the ” China Auto Valley 2021 Intelligent Vehicle Industrial Innovation and Development Forum” in Wuhan. As one of the representatives of simulation manufacturers, Cognata was invited to participate in the panel discussion of Subforum 5: intelligent driving test and evaluation technology. It gathered with representatives of outstanding industry leaders such as Siemens software, PTV, IPG, and Panosim to contribute to jointly promoting the ecosystem of intelligent driving tests and verification!
During the discussion, representative of Siemens, Hanzhi Huang mentioned that many users in the market have not used HIL simulation tools well after purchasing them. He suggested that manufacturers should continue to develop products based on customer needs and common models to make them play a greater value.
Representative of Panosim, Xiangming Li pointed out that on the one hand, this is natural due to the initial running-in of the market. On the other hand, it is also related to different policies. Previously, the automatic driving stage of most enterprises was still in L1-L2, and the simulation requirements mainly lie in regulatory tests and ADAS tests. Most of these cases ran in a limited situation or in a closed zone. However, when it comes to more advanced automatic driving, such as highway simulation, there are dynamic scenes, real traffic flow, and the industry requirements move from regulation and control to decision-making, and from decision-making to perception. It can be predicted that the requirements for simulation will be higher and higher in the future, and the utilization rate will change.
Jerry Wang, the representative of Cognata, also added that in the early stage of demand research and technology matching, the process of technical discussion and demand confirmation between both parties can be strengthened to reduce low utilization due to non-compliance with demand.
In terms of helping users solve specific problems, manufacturers express their views and focus on different aspects, fully demonstrating the richness and diversity of current market demand.
For example, IPG has customers who have more than 30 sensors in a vehicle and need to model all sensors at the same time; PTV has provided users with simulated traffic accident scenes in the traffic collision mitigation project in Japan; Panosim has multiple use cases in multi-V and multi-R high-speed scenarios; Cognata cooperates with ZF on ADAS.AI technology, which reduces the cost by 20%.
No matter what kind of users and needs, as representative of IPG, Xiao Huang said, there will never be a perfect solution, whether software or hardware, SIL or HIL. Everything has to be continuously developed and improved due to the needs of our users.
When asked the ultimate question of “you will die if you don’t develop this one technology, what will you choose?”,
Representative of Panosim, Xiangming Li stressed that simulation, as a paradigm, should first solve the problem of execution. If there is a big difference from the authenticity, the reliability of the simulation is in doubt. I hope to arouse everyone’s continuous thinking and attention.
Jerry Wang, the representative of Cognata, put forward the feasibility of continuous improvement of confidence from the progress of digital twin technology to the new simulation technology based on AI and deep neural networks.
Chang Shen, the representative of PTV, said that parametric simulation has outstanding advantages, yet cloud simulation is the direction that needs to be strengthened in the next step.
Xiao Huang, representative of IPG commented that first, we need to develop multi GPU joint simulation to ensure synchronous output, and secondly, develop ecological partners to ensure effective supply to the whole system.
Experts from China Automobile Center, Baidu Autopilot, Tencent Autopilot, and China Automobile Research Institute also participated in the forum, sharing their insights and outcomes in the industry. Cognata will not forget its original intention and will continue to cultivate deeply, expand and strengthen the new generation of cloud simulation based on AI and deep neural network to better support the needs of local users as always.