Proving Resilient Autonomous Haulage Performance in Extreme Conditions
Executive Summary
EACON Mining Technology's autonomous haulage solution has demonstrated strong resilience, safety and reliability in some of the most challenging mining environments in the world. At Guanghui Energy’s Baishihu Coal Mine and Malang Coal Mine in northwest China, EACON have achieved a record-setting deployment of over 500 autonomous haul trucks at each site. Able to continuously operate in harsh, dusty conditions with wind speeds reaching up to 102 km/h, narrow haul roads and complex terrain, these projects reaffirm the ORCASTRA® platform's scalability, reliability, and OEM-agnostic integration across large mixed fleets.
This case study highlights how EACON's AI-powered 360° perception system, distributed architecture, edge computing, and V2V communication sustain uptime and enable safe, continued operations in poor visibility. By providing real-world evidence of system robustness in extreme weather, dust, heat, and high-stress environments, the deployment underscores EACON’s ability to deliver industry-leading, scalable autonomous haulage solutions that integrate seamlessly with existing mining operations.
Background
Site: Guanghui Energy’s Baishihu & Malang Coal Mines
Location: Northwest China
Commodity: Coal
Fleet: 1000+ factory-fit diesel and hybrid-electric haul trucks manufactured by Tonly and LGMG
Operating Challenges: Dust, strong year-round winds up to 102 km/h, harsh temperatures, narrow load/dump zones, mixed traffic environments
The Challenge
Mining operations at Baishihu and Malang are characterised by abrasive dust and strong winds that can degrade sensor performance and human visibility. Continuous operation in dusty environments raises the stakes for obstacle detection, localisation accuracy and dynamic path planning, especially over long uphill hauls across rugged terrain. At Baishihu, narrow load and dump zones and significant wind gusts (often exceeding 100 km/h) intensify the need for reliable system performance without downtime.
Prior to autonomy, human-driven fleets struggled with safety risks, limited uptime in dust storms, and operational costs tied to labour turnover and hazardous conditions. Stakeholders sought a solution that could sustain high utilisation, uphold safety standards, maintain throughput, reduce labour dependency, and support emission-reduction goals under extreme weather and site constraints, all without requiring major changes to existing mine layouts or haul roads.
The Solution
EACON was selected for its OEM-agnostic ORCASTRA® autonomous haulage platform, proven at scale in real mixed-fleet mining environments.
ORCASTRA® supports diesel, hybrid-electric and battery-electric trucks through retrofit or factory-fit installation, making it adaptable to existing and future fleets. Its three modules (CONDUCTOR, PILOT and CREW) work together to coordinate autonomous mining trucks, conventional vehicles, and equipment across large, complex sites. ORCASTRA® integrates AI-powered 360° perception, distributed intelligence, and multi-sensor fusion (LiDAR, radar, cameras, IMUs and GNSS) to detect, classify and track objects even in severe dust and poor visibility. Real-time LTE/5G and V2V communication enables trucks to share position, speed and task status for coordinated fleet movement.
Able to accurately identify solid hazards or weather-related interference in real time, ORCASTRA® enables safe, continuous operation based on the specific obstacle or environmental conditions encountered. Its distributed architecture allows each vehicle to compute and make decisions independently using edge computing, reducing reliance on central servers and improving resilience during network disruptions. This design minimises single points of failure and supports scalable, plug-and-play fleet expansion. A universally designed drive-by-wire architecture, combined with standardised interfaces between the drive-by-wire layer and the autonomous driving stack, enables rapid integration across multiple OEM truck platforms, with minimal vehicle-specific engineering.
At Baishihu, over 500 autonomous trucks were deployed and aligned with site-specific traffic rules, e.g. left-hand traffic to improve visibility and reduce collision risks. At Malang, the same ORCASTRA® platform was applied across another fleet of 500+ trucks, reinforcing the platform’s repeatability, interoperability, and operational maturity on the world's largest autonomous fleets at single sites.
Results
Continuous operation in dusty, high-wind environments
High uptime and robust obstacle detection in limited visibility
Scalable deployment across 500+ autonomous trucks at multiple sites
Higher operational consistency compared to legacy manned fleets
Reduced on-site safety risks through advanced collision avoidance
Improved traffic flow via customised site rules and intelligent routing
Minimal disruption to existing operations during integration
Cangbao Su, Safety Director at Baishihu Coal Mine, commented on these results: "Since introducing autonomous trucks, the site has seen a clear reduction in safety risks, with mining operations becoming notably more intelligent and efficient. Precision vehicle control and the removal of operators from hazardous zones are helping to create a safer and more streamlined work environment."
Lessons Learned
Tailor safety logic to extreme weather thresholds using AI-powered perception.
Use multi-modal sensing (LiDAR, radar, cameras) to maintain reliability in poor visibility.
Leverage distributed architecture and edge-based computing to reduce network dependency and improve resilience.
Align traffic rules early with stakeholders to maximise safety and efficiency.
Continuous monitoring and iterative perception tuning improve uptime over time.
Drive-by-wire adaptability enables faster deployment across different OEM trucks.
“While many in the OEM-agnostic autonomy space remain focused on small-scale operations, we’ve advanced to large-scale, repeatable deployments." said Elaine Jin, COO of EACON Mining Technology. "Proven in China’s highly complex mining environments, our work with Guanghui demonstrates the product’s maturity and represents a significant step towards intelligent mining, improving safety, efficiency, and reducing risks to human operators in harsh environments.”
Future Outlook
Building on the successful large-scale deployments at Baishihu and Malang, EACON will continue refining ORCASTRA®’s performance in harsh and variable operating conditions. Future efforts will focus on improvements to perception robustness and operational efficiency, informed by real-world data from ongoing operations. These learnings will support repeatable deployment across similarly demanding mine sites, reinforcing EACON’s role as a leading autonomous haulage partner for complex mining environments.
Share this page
