Research Report on AI Foundation Models and Their Applications in Automotive Field, 2024-2025
Research on AI foundation models and automotive applications: reasoning, cost reduction, and explainability
Reasoning capabilities drive up the performance of foundation models.
Since the second half of 2024, foundation model companies inside and outside China have launched their reasoning models, and enhanced the ability of foundation models to handle complex tasks and make decisions independently by using reasoning frameworks like Chain-of-Thought (CoT).
The intensive releases of reasoning models aim to enhance the ability of foundation models to handle complex scenarios and lay the foundation for Agent application. In the automotive industry, improved reasoning capabilities of foundation models can address sore points in AI applications, for example, enhancing the intent recognition of cockpit assistants in complex semantics and improving the accuracy of spatiotemporal prediction in autonomous driving planning and decision.
In 2024, reasoning technologies of mainstream foundation models introduced in vehicles primarily revolved around CoT and its variants (e.g., Tree-of-Thought (ToT), Graph-of-Thought (GoT), Forest-of-Thought (FoT)), and combined with generative models (e.g., diffusion models), knowledge graphs, causal reasoning models, cumulative reasoning, and multimodal reasoning chains in different scenarios.
For example, the Modularized Thinking Language Model (MeTHanol) proposed by Geely allows foundation models to synthesize human thoughts to supervise the hidden layers of LLMs, and generates human-like thinking behaviors, enhances the thinking and reasoning capabilities of large language models, and improves explainability, by adapting to daily conversations and personalized prompts.
In 2025, the focus of reasoning technology will shift to multimodal reasoning. Common training technologies include instruction fine-tuning, multimodal context learning, and multimodal CoT (M-CoT), and are often enabled by combining multimodal fusion alignment and LLM reasoning technologies.
Explainability bridges trust between AI and users.
Before users experience the "usefulness" of AI, they need to trust it. In 2025, the explainability of AI systems therefore becomes a key factor in increasing the user base of automotive AI. This challenge can be addressed by demonstrating long CoT.
The explainability of AI systems can be achieved at three levels: data explainability, model explainability, and post-hoc explainability.
In Li Auto's case, its L3 autonomous driving uses "AI reasoning visualization technology" to intuitively present the thinking process of end-to-end + VLM models, covering the entire process from physical world perception input to driving decision outputted by the foundation model, enhancing users’ trust in intelligent driving systems.
In Li Auto's "AI reasoning visualization technology":
?Attention system displays traffic and environmental information perceived by the vehicle, evaluates the behavior of traffic participants in real-time video streams and uses heatmaps to display evaluated objects.
?End-to-end (E2E) model displays the thinking process behind driving trajectory output. The model thinks about different driving trajectories, presents 10 candidate output results, and finally adopts the most likely output result as the driving path.
?Vision language model (VLM) displays its perception, reasoning, and decision-making processes through dialogue.
Various reasoning models’ dialogue interfaces also employ a long CoT to break down the reasoning process as well. Examples include DeepSeek R1 which during conversations with users, first presents the decision at each node through a CoT and then provides explanations in natural language.
Additionally, most reasoning models, including Zhipu’s GLM-Zero-Preview, Alibaba’s QwQ-32B-Preview, and Skywork 4.0 o1, support demonstration of the long CoT reasoning process.
DeepSeek lowers the barrier to introduction of foundation models in vehicles, enabling both performance improvement and cost reduction.
Does the improvement in reasoning capabilities and overall performance mean higher costs? Not necessarily, as seen with DeepSeek's popularity. In early 2025, OEMs have started connecting to DeepSeek, primarily to enhance the comprehensive capabilities of vehicle foundation models as seen in specific applications.
In fact, before DeepSeek models were launched, OEMs had already been developing and iterating their automotive AI foundation models. In the case of cockpit assistant, some of them had completed the initial construction of cockpit assistant solutions, and connected to cloud foundation model suppliers for trial operation or initially determined suppliers, including cloud service providers like Alibaba Cloud, Tencent Cloud, and Zhipu. They connected to DeepSeek in early 2025, valuing the following:
Strong reasoning performance: for example, the R1 reasoning model is comparable to OpenAI o1, and even excels in mathematical logic.
Lower costs: maintain performance while keeping training and reasoning costs at low levels in the industry.
By connecting to DeepSeek, OEMs can really reduce the costs of hardware procurement, model training, and maintenance, and also maintain performance, when deploying intelligent driving and cockpit assistants:
Low computing overhead technologies facilitate high-level autonomous driving and technological equality, which means high performance models can be deployed on low-compute automotive chips (e.g., edge computing unit), reducing reliance on expensive GPUs. Combined with DualPipe algorithm and FP8 mixed precision training, these technologies optimize computing power utilization, allowing mid- and low-end vehicles to deploy high-level cockpit and autonomous driving features, accelerating the popularization of intelligent cockpits.
Enhance real-time performance. In driving environments, autonomous driving systems need to process large amounts of sensor data in real time, and cockpit assistants need to respond quickly to user commands, while vehicle computing resources are limited. With lower computing overhead, DeepSeek enables faster processing of sensor data, more efficient use of computing power of intelligent driving chips (DeepSeek realizes 90% utilization of NVIDIA A100 chips during server-side training), and lower latency (e.g., on the Qualcomm 8650 platform, with computing power of 100TOPS, DeepSeek reduces the inference response time from 20 milliseconds to 9-10 milliseconds). In intelligent driving systems, it can ensure that driving decisions are timely and accurate, improving driving safety and user experience. In cockpit systems, it helps cockpit assistants to quickly respond to user voice commands, achieving smooth human-computer interaction.
Research on Zonal Architecture: Smart Actuators (Micro-motors) and Application Trends in Sub-scenarios, 2026
Smart Actuator and Micro-motor Research: Under Zonal Architecture, Actuators Are Developing towards Edge Computing, 48V, and Brushless Motors.
The core components of automotive zonal architecture mai...
China Passenger Car Navigate on Autopilot (NOA) Industry Report, 2025
In 2025, NOA standardization was popularized, refined and deepened in parallel. In 2026, core variables will be added to the competitive landscape.
The evolution of autonomous driving follows a clear...
Smart Car OTA Industry Report, 2025-2026
Automotive OTA Research: In the Era of Mandatory Standards, OTA Transforms from a "Function Channel" to a New Stage of "Full Lifecycle Management"
Driven by the development and promotion of AI and so...
Automotive AI Box Research Report, 2026
Automotive AI Box Research: A new path of edge AI accelerates
This report studies the current application status of automotive AI Box from the aspects of scenario demand, product configuration, and i...
Automotive Fragrance and Air Conditioning System Research Report, 2025
Automotive Fragrance and Air Purification Research: Intelligent Fragrance Equipment to Exceed 4 Million Units by 2030, "All-in-One" Integrated Purification Becomes Mainstream
The "Automotive Fragranc...
Intelligent Vehicle Cockpit-driving Integration (Cockpit-driving-parking) Industry Report, 2025
Cockpit-Driving Integration Research: 36% CAGR by 2030, Single-Chip Cockpit-driving integration Solutions Enter Mass Production
ResearchInChina releases the "Intelligent Vehicle Cockpit-driving Integ...
Research Report on Overseas Layout of Chinese Passenger Car OEMs and Supply Chain Companies, 2025
Automotive Overseas Expansion Research: Accelerated Release of OEM Overseas Production Capacity, Chinese Intelligent Supply Chain Goes Global
This report conducts an in-depth analysis of the current ...
Passenger Car Intelligent Steering Industry Research Report, 2025-2026
Intelligent steering research: Rear-wheel steering prices drop to RMB200,000-250,000
1. Rear-wheel steering installations increased by 36.5% year-on-year.
From January to October 2025, the number of...
Global Autonomous Driving Policies & Regulations and Automotive Market Access Research Report, 2025-2026
Research on Intelligent Driving Regulations and Market Access: New Energy Vehicle Exports Double, and "Region-Specific Policies" Adapt to Regulatory Requirements of Various Countries in A Refined Mann...
Two-wheeler Intelligence and Industry Chain Research Report, 2025-2026
Two-Wheeler Electric Vehicle Research: New National Standard Drives Intelligent Popularization, AI Agent Makes Its Way onto Vehicles
ResearchInChina releases the "Two-wheeler Intelligence and Industr...
China Smart Door and Electric Tailgate Market Research Report, 2025
Smart Door Research: Driven by Automatic Doors, Knock-Knock Door Opening, etc., the Market Will Be Worth Over RMB100 Billion in 2030.
This report analyzes and researches the installation, market size...
New Energy Vehicle Thermal Management System Industry Research Report, 2025-2026
Policy and Regulation Drive: Promoting the Development of Electric Vehicle Thermal Management Systems towards Environmental Compliance, Active Safety Protection, and Thermal Runaway Management
Accord...
Intelligent Vehicle Redundant Architecture Design and ADAS Redundancy Strategy Research Report, 2025-2026
Research on Redundant Systems: Septuple Redundancy Architecture Empowers High-Level Intelligent Driving, and New Products Such as Corner Modules and Collision Unlock Modules Will Be Equipped on Vehicl...
Passenger Car Mobile Phone Wireless Charging Research Report, 2025
Automotive Wireless Charging Research: Domestic Installation Rate Will Exceed 50%, and Overseas Demand Emerges as Second Growth Driver.
The Passenger Car Mobile Phone Wireless Charging Research Repor...
Automotive 4D Radar Industry Research Report 2025
4D radar research: From "optional" to "essential," 4D radar's share will exceed 50% by 2030.
1. 4D imaging radar has transformed from an "optional" to a "must-have" sensor.
4D radar adds the detecti...
China Automotive Multimodal Interaction Development Research Report, 2025
Research on Automotive Multimodal Interaction: The Interaction Evolution of L1~L4 Cockpits
ResearchInChina has released the "China Automotive Multimodal Interaction Development Research Report, 2025"...
Automotive Vision Industry Report, 2025
Automotive Vision Research: Average Camera Installation per Vehicle Reaches 5.2 Units, and Front-View Tricam Installation Exceeds 1.2 Million Sets.
From January to September 2025, the total installa...
Automotive Infrared Night Vision System Research Report, 2025
Automotive night vision research: The rise of infrared AEB, with automotive infrared night vision experiencing a 384.7% year-on-year increase from January to September.
From January to September 2025...