The fleet has exceeded 3,100 units, with a market share surpassing 53%: EACON (07687), the global leader in mining physical AI commercialization, presents its mid-year report.

Focusing solely on the surface-level data of "the number of operating vehicles nearly doubling" can easily lead to an underestimation of the deeper value of EACON (07687) mid-term performance in 2026. There are three core propositions that the capital market really needs to clarify: Has the unmanned mining truck industry crossed the commercialization tipping point? Can over 3,100 operating vehicles build a quantifiable scale barrier? How will fleet expansion, asset-light transformation, and breakthroughs in overseas and metal mining scenarios gross margin, profit quality, and long-term growth potential? As of June 30, 2026, the company has over 3,100 autonomous mining trucks in operation, almost double the number from the same period in 2025, with more than 1,000 pending delivery orders. According to industry data cited in the interim report, China has approximately 6,000 operating unmanned mining trucks, and the companys market share exceeds 53%. This is by no means merely a result of solution delivery. In the production-oriented mining scenario where Siasun Robot & Automation has first achieved scalable implementation, EACON has built the largest physical AI operation cluster in the country. More critically, scale expansion, lightweight models, high-end mining types, and market globalization are advancing simultaneously, collectively solidifying the core support for becoming the "global leader in physical AI commercialization for mining": verifying commercialization maturity through operational scale, validating AI capability levels through complex working conditions, and confirming replicability across scenarios and regions. 1. From 12% to 50%+: Industry has crossed the commercialization tipping point According to Frost & Sullivan data, the market size for unmanned mining solutions in China is expected to reach 3.8 billion yuan by 2025, with a compound annual growth rate (CAGR) of 171.4% from 2021 to 2025; during the same period, the penetration rate of unmanned mining truck sales is about 12%, which is expected to exceed 50% by 2030. This indicates that the industry has completed demonstration verification and officially entered the stage of scaled procurement and normalized operation. The interim report shows that in the first half of 2026, approximately 1,500 new autonomous mining trucks came into operation in China, while the penetration rate continues to accelerate. For mining companies, automation has evolved from a technology showcase to a necessity solution that addresses safety, efficiency, labor, and lifecycle costs. The global market also presents vast opportunities. Frost & Sullivan predicts that the global unmanned mining solutions market will grow from approximately $1 billion in 2025 to around $7.3 billion in 2030, with a CAGR of 47.4%; among them, the Australian market is expected to exceed $2.9 billion by 2030. The rapid expansion of the domestic market combined with the ongoing growth in high-value overseas markets has laid a solid industrial foundation for EACON to transition from a domestic leader to a global pioneer. Policies continue to accelerate the industry. The revised "Coal Mine Safety Regulations" came into effect in February 2026; the National Mine Safety Administration and the Ministry of Industry and Information Technology initiated pilot projects for Siasun Robot & Automation applications, focusing on unmanned transportation in open-pit mines; related safety production policies constantly push large mines to accelerate automation and unmanned upgrades. The mining scenario is naturally well-suited for physical AI implementation: operational routes are relatively fixed, transportation tasks are repeated frequently, and systems need to complete the entire closed-loop of perception, decision-making, and execution in real environments. Unlike laboratory demonstration AI, unmanned mining trucks must withstand real-world tests involving all-weather production, complex traffic organization, and extreme working conditions. Referring to the commercialization rules of general Siasun Robot & Automation industries, a company's core competitiveness has never been a one-off technology demonstration but rather the ability to consolidate algorithms into reproducible products, deliver them in volume to real scenarios, and achieve continuous iterations through operational data. EACON has implemented this logic in the mining transportation sector, charting a path for the large-scale application of physical AI in mining. 2. From Vehicle Scale to AI Clusters: Building Scalable System Capabilities What truly builds a barrier is not just the total number of vehicles, but the system capabilities accumulated through high-density fleet operations. As of the end of the reporting period, the companys solutions have covered 38 mining sites, adding 12 new ones compared to the same period last year; among them, there are nine sites with fleets of over 100 trucks, and four sites with fleets of over 200 trucks, with the largest single fleet reaching 566 trucks. A fleet of over 100 trucks is not merely a simple accumulation of individual vehicles. The larger the fleet, the higher the requirements for high-concurrency dispatching, the coordination of the entire loading and unloading process, anomaly management, safety redundancy, remote operation and maintenance, and coordination between manned and unmanned equipment. The long-tail scenario data generated from normalized operations will continuously feed back into algorithm iterations, simulation testing, delivery standards, and operational maintenance systems, eventually forming a virtuous cycle of scale expansion data accumulation algorithm evolution delivery efficiency improvement further scale expansion. Currently, the companys solution has achieved large-scale implementation in complex environments, such as altitudes over 5,000 meters, ultra-low temperatures of minus 40C, extreme heat, rain, fog, dust, and scenarios with high traffic density. During the reporting period, key breakthroughs were made in adapting to narrow roads in deep opencast mines, deploying in weak network and weak positioning environments, and adapting to metal mining processes: the width requirement for road adaptation has been reduced by over 20%, and unmanned driving deployment can be completed based on public 4G networks. This also means that EACONs barrier is not merely leading in the number of vehicles, but rather a comprehensive capability barrier built upon operational scale operational duration scenario complexity. The asset-light transformation further amplifies the capital efficiency of this capability. As of the end of the reporting period, among the over 3,100 operating vehicles, the company provided about 600 in fleet mode, while customers provided about 2,500 in fleet mode; based on revenue, the proportion of customer-provided fleet mode increased from 62.7% in 2025 to 65.6% in 2026. The differences in profitability have become evident: in the first half of 2026, the gross margin of the customer-provided fleet mode reached 32.3%, significantly higher than the companys provided fleet mode. Meanwhile, the company's capital expenditures dropped from 538 million yuan in the same period last year to 117 million yuan, a year-on-year decrease of 78.3%, primarily due to a significant reduction in spending on unmanned mining truck purchases. The essence of this transformation is that the mining companies and engineering firms take on vehicle assets and scene resources, while EACON focuses on delivering algorithms, sensor systems, complete vehicle control, dispatching platforms, and ongoing maintenance services. The companys value focus is shifting from "owning vehicle assets" to "operating intelligent fleets, accumulating scenario data, and delivering intelligent systems." The platform capability is also reflected in vehicle model adaptation. As of the end of the reporting period, the companys Yushi line control platform has been adapted to over 100 models from 16 main engine manufacturers, covering six rigid mining trucks from three major domestic and foreign manufacturers. Leveraging a unified control and algorithm framework, existing scene data and model capabilities can be quickly transferred to new models, significantly reducing the redundant development costs of cross-model deployment. 3. From Scale Expansion to Value Upgrade: Multi-Dimensional Verification of the Commercialization Flywheel The mid-term performance of 2026 shows that the company achieved revenue of 549 million yuan, a year-on-year decline of 17.3%; gross profit of 104 million yuan, a YOY increase of 844.3%; gross margin jumped from 1.7% in the same period last year to 18.9%; the loss for the period reached 189 million yuan, a year-on-year reduction of 20.2%; and the adjusted loss was 168 million yuan, narrowed by 22.2% year-on-year. The short-term pressure on revenue is a phase phenomenon resulting from the combined effects of strategic transformation and delivery rhythm: first, the company proactively disposed of early self-operated vehicles, shrinking capital-intensive operations and focusing on high value-added technical services; second, some customer fleet projects fall within the delivery cycle, with only partial vehicle acceptance completed, and the remaining revenue will be gradually recognized as the projects progress. Meanwhile, revenue from mining digital solutions has increased by more than 10 times year on year, indicating that the business boundaries are extending from single unmanned transportation to complete scene digitalization involving production scheduling, equipment management, and more. The significant recovery in gross margin is a direct validation of the effectiveness of the asset-light transformation. The proportion of the high-margin customer-provided fleet mode has continuously increased, significantly optimizing the overall profitability structure. Future market tracking can closely monitor indicators such as the proportion of customer fleet mode, single-vehicle operating revenue, project delivery cycles, accounts receivable turnover, and operating cash flow to observe the efficiency of the transmission from scale expansion to profitability quality. Breakthroughs in high-value markets further open up the value ceiling for individual vehicles and projects. In the overseas market, the company has made substantial progress in its Australian project in cooperation with Zijin Mining Group's Norton Gold Fields and global mining service provider Thiess: the first batch of six mining trucks equipped with the companys unmanned system has entered real production at the Havanna open-pit gold mine in the Kalgoorlie mining area of Western Australia. This marks the first unmanned mining truck project to enter a real production environment in the Kalgoorlie gold mining area, making the company the first third-party unmanned driving solution provider to establish operations in Australia. In the domestic high-value scenarios, the company deployed 60 unmanned mining trucks at Zijin Mining Groups Julong copper mine in May 2026, creating the largest unmanned driving project by single mine scale in the metal mining industry. As of the end of the reporting period, the company has reached five domestic metal mining service projects, covering core mines such as the Julong copper mine and the Zijinshan gold-copper mine. The processes in metal mining are more complex, and the Australian market imposes higher requirements for compliance systems, service responses, and localized operations. The successful implementation of these two types of scenarios indicates that the company has completed the capability leap from a single coal mine context to complex mining types and markets in developed overseas countries. Technological barriers are also being solidified concurrently. The company participated in the completion of the project Key Technologies and Applications for Green, Intelligent, Safe, and Efficient Mining in Large Open-Pit Mines, which won the second prize of the National Scientific and Technological Progress Award in 2025, marking the first time this award has been granted to an unmanned driving project; the self-developed IMRv2 algorithm ranked first in the interactive prediction and trajectory prediction leaderboard of the Waymo Open Motion Dataset. As of the end of the reporting period, the company holds 284 authorized patents in China and has filed 44 PCT international patents. From a capital market perspective, the next stage of value verification for the company will upgrade from fleet scale to four dimensions: first, can the revenue share and profit level of the customer fleet model continue to improve? Second, can the operating efficiency of individual vehicles and projects continue to enhance? Third, can delivery cycles, payment efficiency, and operating cash flow continue to be optimized through platform replication? Fourth, can high-value scenarios in Australia and metal mining be replicated in volume? This is exactly the core difference between the "global leader in physical AI commercialization for mining" and ordinary self-driving suppliers: their competitiveness lies not in one-off technology deployment but in the sustained delivery, continuous operation, and ongoing value creation capabilities across different mining types, regions, and regulatory environments. In summary, the core value of this mid-term performance lies not merely in the numerical increase in fleet size, but in a complete closed-loop commercialization logic: backed by an industry penetration rate that continues to rise, the company constructs barriers of data and delivery with its operating fleet of over 3,100 vehicles; through asset-light transformation, it shifts vehicle capital expenditures to ecological partners, anchoring the value focus on algorithms, platforms, and services; and with breakthroughs in metal mining and Australian projects, it verifies the replicability in high-value markets. If general Siasun Robot & Automation enterprises are pushing the industry from functional validation to mass production delivery, then EACON has already led mining Siasun Robot & Automation into the new stage of fleet clustering, operational platformization, and market globalization. This serves as the most solid reality underpinning its positioning as the "global leader in physical AI commercialization for mining."