ResearchInChina released the "OEMs and Tier 1 Suppliers' Cost Reduction and Efficiency Enhancement Strategy Analysis Report, 2025", summarizing hundreds of cost reduction strategies to provide reference for automotive industry practitioners.

Global Automotive Industry Layoff Plans Exceed 200,000, Highlighting Urgent Need for Cost Reduction

In 2025, the cold winter in automotive industry continues, with frequent news of enterprise closures, bankruptcies, and dissolutions:
In January 2025, HYCAN Auto suspended operations and was taken over by its parent company GAC Group for subsequent disposal.
In May 2025, Neta Auto was applied for bankruptcy review and subsequently entered bankruptcy proceedings.
In July 2025, GAC Fiat Chrysler was formally ruled bankrupt by the court due to no possibility of reorganization, and its core assets failed to sell at auction multiple times.
In November 2025, Haomo.AI suddenly notified all employees of a work suspension, with accounts frozen, wages and social security payments halted, and operations basically suspended.

Accompanying this is widespread layoff information from enterprises. Over the past year, the global automotive industry chain has seen frequent layoffs, with companies such as Mercedes-Benz, Audi, Volkswagen, Ford, Nissan, ZF Friedrichshafen, and Bosch successively announcing large-scale layoff plans. According to incomplete statistics, the number of planned layoffs by OEMs exceeds 120,000, and that by leading supplier giants exceeds 100,000.

At the same time, declining profit margins caused by price wars and rising R&D costs driven by intelligence have continued to increase pressure on OEMs and suppliers, making the urgency of cost reduction and efficiency enhancement more prominent.

I. Five Core Strategies for Supply Chain Cost Reduction

ResearchInChina's "OEMs and Tier 1 Suppliers' Cost Reduction and Efficiency Enhancement Strategy Analysis Report, 2025" divides enterprise cost reduction strategies into two categories:
The first category is management-driven cost reduction, mainly including supply chain optimization and vertical integration, organizational and process optimization, marketing and sales innovation, circular economy and compliance management;
The second category is technology-driven cost reduction, mainly including intelligent driving technology cost reduction, intelligent cockpit technology cost reduction, powertrain and chassis technology cost reduction, and other related technology cost reduction.

In supply chain cost reduction, OEMs often expand their control over the supply chain through vertical integration, component standardization, and transparent supply chains; suppliers often expand their business volume in the supply chain and improve profit space through modular supply and joint R&D.

Supply Chain Cost Reduction Strategy 1: Vertical Integration

Supply chain vertical integration refers to a management strategy in which enterprises achieve a high degree of control over the entire value chain by controlling multiple consecutive links in the supply chain (from raw material procurement, production and manufacturing to sales and distribution). Its core is to reduce dependence on external suppliers and improve efficiency and competitiveness through internalization of key links.

Representative Enterprise 1: BYD

BYD's vertical integration capability has become an industry benchmark, covering key links such as raw materials (lithium ore resources), core components (IGBT, cathode materials, electrolyte, separators), and core auto parts (batteries, electric drives, chips, automotive electronics), forming a vertical integration model from "ore to vehicle".

Taking BYD Seal as an example, according to a disassembly report by UBS, self-produced components account for approximately 75% of the vehicle. The high degree of supply chain integration has enabled BYD to significantly reduce component procurement costs while enhancing supply chain safety and stability.

Representative Enterprise 2: Leapmotor

Leapmotor independently develops core components such as three-electric system (battery, motor, ECU), intelligent driving, cockpit system, electric drives, batteries, and EE architecture, covering 70% of the vehicle BOM cost. For example, the electric drive assembly, battery PACK, and intelligent cockpit are all independently developed and produced, reducing the R&D cost per vehicle by 40% compared with the outsourcing mode.

Supply Chain Cost Reduction Strategy 2: Modular Supply

Modular supply refers to a supply chain model in which automakers decompose automotive products into independent functional modules, realize the combination between modules through standardized interfaces, and professional suppliers are responsible for module development, production, and integration. This not only simplifies supply chain management but also reduces costs through bulk procurement. For suppliers, although the profit margin of individual components may be compressed, they can gain more market opportunities by providing overall solutions.

Supply Chain Cost Reduction Strategy 3: Component Standardization

Core of automotive modularization is standardization, such as ensuring module compatibility through standardized interfaces. NIO found that interfaces of various components on the vehicle, such as seats, air conditioners, and reading lights, were different, resulting in high matching costs. Therefore, the R&D team standardized component interfaces, optimizing from hundreds of interfaces to about 40-50. After reducing the types of supplies, the cost per vehicle decreased from 2,000 yuan to about 1,000 yuan.

Supply Chain Cost Reduction Strategy 4: OEMs & Suppliers Joint Cost Reduction

In the battle of cost reduction, while OEMs and suppliers can achieve certain results by acting alone, the effect may be more prominent if both parties cooperate closely. By establishing a close cooperative relationship, both parties can share resources, technologies, and market information, thereby achieving more efficient collaborative innovation and cost optimization.

For example, OEMs and suppliers can reduce technology development costs through joint R&D. OEMs usually have a deeper understanding of market demand and product positioning, while suppliers have professional advantages in component design and manufacturing. Both parties can jointly formulate technology roadmaps, concentrate resources to tackle key technical problems, and avoid duplicate investment and resource waste.

Supply Chain Cost Reduction Strategy 5: Transparent Supply Chain

A transparent supply chain refers to the realization of real-time, credible, and two-way sharing of core data such as production, quality, logistics, inventory, and production capacity between OEMs and partners with the help of digital technology. Its core is to break industry "data silos", realize information transparency and sharing between upstream and downstream of the supply chain, and promote real-time and efficient collaboration between enterprises.

On October 22, 2025, the Transparent Supply Chain Ecosystem-Auto (TSCE-Auto) was officially launched. Initiated by the China Society of Automotive Engineers and Shanghai International Automobile City, it was jointly established by 18 major domestic passenger car enterprises. OEMs include: FAW Group, Dongfeng Motor Corporation, Changan Automobile, GAC Group, BAIC Group, JAC Group, SAIC-GM-Wuling, Geely Auto, Chery Automobile, BYD Auto, NIO, XPeng Motors, Li Auto, Leapmotor, Seres, and Xiaomi Auto.

NIO has taken the lead in joining the transparent supply chain system. As of October 2025, 63 suppliers have accessed NIO's transparent supply chain system, with plans to cover 16 scenarios, 9 of which have been implemented, including real-time transparency and early warning of production quality, end-to-end inventory transparency, automated order fulfillment, and production capacity utilization calculation.

II. Five Core Strategies for Intelligent Driving System Cost Reduction

According to the price distribution of mainstream models in the Chinese market, Zhuoyu Technology believes that the cost range of L2+ intelligent driving systems is 5,000 yuan to 15,000 yuan. Momenta proposed the Moore's Law for intelligent driving, believing that hardware costs will be halved every two years. For example, for map-free urban NOA, the cost reduction path of intelligent driving domain controllers is: 8,000￥ → 4,000￥ → 2,000￥.

Currently, cost reduction for intelligent driving systems mainly focuses on system cost reduction, sensor cost reduction, test cost reduction, etc. System cost reduction is mainly achieved through vision-only solutions, map-free intelligent driving, integrated intelligent driving, and localized replacement; sensor cost reduction is mainly achieved by replacing traditional high-cost sensors with 4D radar, low-cost lidar, binocular/trinocular cameras, laser-vision systems, RGB-IR cameras, etc.; test cost reduction is achieved by improving simulation test efficiency through AI, world models, and automation.

Intelligent Driving Cost Reduction Strategy 1: Vision-Only Solution

The vision-only intelligent driving solution is a technical route that uses cameras as the main perception hardware to achieve environmental perception, decision, and control through algorithms and neural networks. With advantages such as low cost and fast iteration, the vision-only solution has become an important path for the popularization of intelligent driving. Tesla, XPeng, BYD, Huawei, etc. have all launched intelligent driving solutions based on vision-only. Among them, Tesla relies entirely on cameras, while XPeng, BYD, etc. take cameras as the core, supplemented by millimeter-wave radar or ultrasonic radar to enhance perception.

Intelligent Driving Cost Reduction Strategy 2: Integrated Intelligent Driving

Cost reduction through integration of intelligent driving and intelligent cockpit refers to a strategy of integrating intelligent driving functions and intelligent cockpit functions into a single domain controller to achieve centralized function control. Integrated intelligent driving can achieve system cost reduction through sensor sharing, a single SoC, and simplified interfaces, mainly including three directions: driving-parking integration, cockpit-parking integration, and cockpit-driving integration. The development of E/E architecture from distributed to domain controller architecture provides the possibility for it.

Case 1: Yihang.AI
Yihang.AI's single-chip Lite version of integrated driving and parking domain controller only needs to develop a set of basic software and middleware. Through technological innovations such as algorithm pruning, knowledge distillation, and shared backbone, the algorithm computing power is optimized, and the cost of the entire domain controller is controlled at around 1,000 yuan.

Case 2: NavInfo
NavInfo's driving-parking integrated solution PhiGo Entry developed based on Horizon Journey 6B has a cost as low as the "thousand-yuan level", has been designated by leading international automakers, and is planned to be mass-produced in Q2 2026.

Intelligent Driving Cost Reduction Strategy 3: Localized Replacement

Localized replacement of automotive components can significantly reduce costs, and the same applies to intelligent driving systems. Rapid development of local Chinese chip suppliers, sensor suppliers, and system integrators has continuously expanded the space for localized replacement of intelligent driving systems.

Taking intelligent driving chips as an example, the product capabilities of local enterprises such as Horizon Robotics, Huawei, and Black Sesame Technologies have been continuously improved. Especially in the field of medium and low computing power intelligent driving, they have formed a replacement effect on NVIDIA, Qualcomm, etc.

Low computing power chips refer to chips with computing power below 30TOPS, mainly used in lightweight driving and parking systems, such as low-level assisted driving functions such as reverse imaging, forward collision warning, lane keeping assistance, and basic automatic parking. In the low computing power field, domestic intelligent driving chip manufacturers have competitive advantages no less than traditional leaders, with a high and stable localization replacement rate. Current domestic solutions are mainly based on Horizon J2, J3, 6B, and Black Sesame A1000L.

Medium computing power chips refer to chips with computing power between 30 and 150TOPS, which can support complex autonomous driving functions, such as Highway NOA, Urban Memory NOA, and Memory Parking. Some models may also provide Urban NOA functions. In the medium computing power chip market, domestic chips such as Horizon J5/6E/6M and Black Sesame A1000 have formed replacements for overseas chips such as TI TDA4, NVIDIA Xavier, Orin N, EyeQ6H, and EyeQ7.

High computing power chips refer to chips with computing power exceeding 150TOPS, mainly for high-level assisted driving and even autonomous driving scenarios. Currently, the shipped products on the market mainly include NVIDIA Orin X, Tesla FSD, and Huawei Ascend 610.

Intelligent Driving Cost Reduction Strategy 4: 4D Radar

4D radar is an upgraded form of traditional millimeter-wave radar. "4D" refers to its ability to simultaneously obtain four-dimensional information of the target: distance, speed, horizontal angle, and vertical height. In terms of performance, 4D radar can break through limitations in static target recognition, lateral movement detection, height recognition, distinguishing adjacent objects, and detecting hidden vehicles. Compared with lidar, it is very close in certain performance indicators, and its performance is less affected in harsh weather and environments, with lower cost than lidar.

The design of 4D radar chips mainly includes cascaded solutions and single-chip solutions. With advantages in integration, cost, and size, the single-chip solution has become the future direction of 4D radar and an ideal replacement solution in the cost reduction of intelligent driving sensors.
Integration: The single-chip solution replaces the traditional multi-chip cascaded solution (e.g., 12-transmitter and 16-receiver require 4 MMICs), reducing volume by more than 50% and power consumption by 30-70%.
Cost: Calterah CAL77S344 has a unit price reduced to the hundred-yuan level; ANDAR ADT6220A (60GHz AiP chip) adopts FCCSP packaging to further reduce module costs.
NXP predicts that single-chip will become the mainstream solution for corner radars after 2025.

Intelligent Driving Cost Reduction Strategy 5: LiDAR-Vision System

The LiDAR-vision system is an innovative sensor design jointly developed by Zhuoyu Technology and Fuyao Glass, integrating lidar, front-facing trinocular camera, and inertial measurement unit into a single module, which can achieve higher-precision "spatiotemporal synchronization" between LiDAR and camera.

Compared with current mainstream "lidar + front-facing camera" solution on the market, Zhuoyu's LiDAR-vision solution can reduce costs and achieve 100% function realization and performance replacement under the premise of effectively simplifying the cost of vehicle integration, layout, and production calibration.