China is leading the world in mine-vehicle automation outside of Australia and may be able to surpass its southern counterpart, in terms of total fleet population, this year. InvestMETS.com spoke with EACON Mining Technology co-founder and chair, Lei Zhang, after we recently quizzed former Anglo American head of innovation, David Pugh, on his “system level autonomy” blog.
Pugh said vehicle-level autonomy in mining had demonstrated safety and productivity benefits.
But he said, “this value is dwarfed by system and fleet-level autonomy”.
Asked how far away he thought mine “system-level” autonomy was, Pugh said: “From a technical perspective, I do think the industry could be there within two years, but I think it will likely take a lot longer for a full deployment, because the main challenges include integrating the diverse range of equipment and systems on a site, and the people and process changes needed.”
Based on information currently circulating in the public domain, we suggested to EACON’s Lei Zhang that rapid progress in China appeared to indicate the industry there was ready to widely deploy higher levels of autonomy at brownfield, not just greenfield, mines.
We also asked about autonomous loading machines, interoperability (system) challenges and how and why there had been such a rapid development and uptake of heavy industry auto-mobility in China.
Lei Zhang: It will take some time to have comprehensive deployment of autonomous driving, including mining trucks, auxiliary operation equipment, and loading-vehicle autonomous driving.
EACON’s partner has achieved remote control driving of excavators, but the loading efficiency is difficult to reach manual level. Indeed, some companies in the industry now are dedicated to autonomous driving of mining equipment other than mining trucks, but I estimate it will be very challenging.
For example, for excavators, autonomous operations are relatively easy to achieve for topsoil stripping operations. But ore-loading operations require automated ore identification and corresponding behaviours due to the different grades and boundaries of the ore.
The development of general artificial intelligence technology in the future will be key to improving capabilities in perception and decision-making, which may take some time.
China’s mines are developing whole-mine-level production management systems.
At the TBEA South Pit mine EACON vehicles account for nearly 50% of the total mine fleet and we are helping the mining company to establish a whole-mine-level production management system. EACON fleet data has been well monitored in real-time. Non-EACON vehicles have also contributed some location and production-related near real-time data. To achieve the above information sharing, it is necessary for the miner to coordinate between EACON and other fleet operators.
In Australia, autonomous driving systems are mainly provided by OEMs. The OEM only provides a limited number of vehicle models. Therefore, if a mine uses trucks from different brands and models, it may be difficult to achieve autonomous driving on the whole-mine-level for all trucks.
Of course, there are also suppliers providing OEM-agnostic retrofit autonomous driving solutions. However, due to different brand models, different wire control systems need to be developed and debugged. Although it can achieve generalisation to a certain extent, the speed of landing and verification may be slightly slower.
Still, this is a very valuable direction so EACON is also providing these retrofit autonomous driving solutions.
In China, EACON is developing business fast and benefits from the open co-operation mentality of OEMs.
EACON provides all or part of the drive-by-wire control design and OEMs open up their control protocols. The EACON autonomous driving system can be accurately and quickly implemented on different OEM models. For example, in the TBEA South pit-mine, EACON’s autonomous driving fleet includes four models from three OEM brands. On some truck models, EACON not only designs the drive-by-wire control system, but also battery/hybrid electric energy systems.
In addition to OEM co-operation and support, our understanding of working scenarios and our ability to iterate in research and development contribute to our ability to achieve large-scale autonomous driving.
When EACON was first implemented for autonomous driving it also required a dedicated work area for autonomous trucks to carry out operations.
Now we have the ability to mix manned and autonomous trucks for full operation, which will help us achieve whole-mine-level autonomous operation more quickly.
Ramping up autonomous driving at greenfield and brownfield takes a series of steps.
During this period it is better to let manned and autonomous trucks work together to secure production efficiency. Of course if a mining company is willing to leave some time for debugging and operation it can also adopt a pure autonomous driving fleet from the beginning.
InvestMETS: How do you see the systems integration/interoperability challenge? Is it less of a barrier at the mines you have deployed systems to in China? Why or why not?
Lei Zhang: The challenges encountered in China will be different from those in Australia.
Since we directly provide factory-fitted autonomous driving solutions through cooperation with OEMs, we don’t have to worry about integration and interoperability issues after the system is deployed to the mine.
However, before the production of autonomous driving vehicles, establishing a cooperative relationship with OEMs, jointly building the vehicle’s drive-by-wire control capabilities, and coordinating control protocols, all require a lot of effort.
The challenges of system integration/interoperability when deployed in mines mainly lie in connecting with infrastructure, such as traffic lights.
Deployment of truck dispatch systems in China is not as wide as in Australia. Currently, we are deploying our own autonomous truck dispatch systems and will also help mining companies establish management systems for both EACON and non-EACON fleets throughout the mine.
“The next 1-2 years are likely to be the expansion period for the implementation of autonomous driving in China”
EACON has accumulated a lot of relevant experience in the TBEA South pit-mine, and the TBEA South pit-mine has also become a very good model in China market. Some other mine owners highly recognize the successful experience of TBEA South Pit mine, so we can quickly push forward in system integration/interoperability.
InvestMETS: How would you characterise the receptiveness in the mining market in China, and the appetite, for fully autonomous fleet/system adoption?
Lei Zhang: We believe that the acceptance of fully autonomous driving fleets and systems in the Chinese mining industry market is rapidly increasing, and the demand is also constantly increasing. The next 1-2 years are likely to be the expansion period for the implementation of autonomous driving in China.
With the increase in labour costs the economic value advantage of manual labour is gradually decreasing.
The cost of using an autonomous vehicle fleet/system will transition to a flat or even lower level than labour costs in the future.
At the same time, considering various factors such as safety, relevant government regulatory departments are gradually releasing new policies and regulations, putting forward higher requirements for the deployment scale of autonomous driving systems in the future mining industry. As a result, mining companies are expanding their demand for mature autonomous driving fleets/systems. Companies are even willing to pay more for autonomous driving than for human labour after learning that it is safer.
The increasingly mature and stable autonomous driving fleet/system is also increasing market acceptance of it.
EACON’s technology has played a key role in normalising AHS [autonomous haulage system] technologies in the Chinese mining industry.
After three years of constant system and operational improvement, the TBEA South Pit has become a benchmark case in the industry, and many customers have greatly reduced their concerns about autonomous driving after visiting.
In terms of fleet size and output we account for almost 50% at the TBEA South Pit. The safety, stability and efficiency of the entire fleet and system have been verified by actual production.
InvestMETS: Circling back to autonomous shovels and loaders, is there an appetite to automate loading machines further, or at least remote control them?
Lei Zhang: I don’t think fully autonomous loading is likely to happen anytime soon. The main difficulty is caused by non-standard material specifications, and the system needs to have a deeper semantic understanding of the loading scenario of non-standard materials.
There are two ways to develop autonomous loading in the future. On one hand is the significant improvement in general artificial intelligence, which brings about an improvement in perception and decision-making levels. The system needs to be able to better handle material recognition and operational decision-making problems.
On the other hand, create a unified and standardised environment for autonomous loading systems through standardised operation. For example, follow certain processes and regulations during the blasting operation to provide conditions for standardised loading. By standardising excavator operations, the difficulty of achieving full-automation can be reduced.
However, this approach comes with additional costs.
InvestMETS: A highly experienced mining-tech advisor said to me recently he thought mining was getting closer to its “iPhone moment”. This is about the convergence and integration of a host of platform, or enabling, technologies reaching a stage where the truly game-changing technology or system emerges and, potentially, disrupts.
What is your view on the smart-phone analogy?
Lei Zhang: The metaphor is very vivid and well illustrates the current state of mining autonomous driving. Key technologies are constantly emerging and upgrading, and the industry is ushering in a moment of change.
Firstly, there have been changes in energy technology, evolving from traditional diesel power to zero emissions and upgrading to hybrid, pure electric, hydrogen-powered and other driving methods. While collaborating with OEMs in the design of new energy systems we also offer drive-by-wire controlled designs, which enable new energy vehicles to possess the underlying capabilities of autonomous driving.
Furthermore, the development of artificial intelligence related technologies has made autonomous driving technology increasingly powerful. Companies like EACON have delved into production scenarios and promoted the practical application of technology.
The progress of infrastructure is also enormous. Currently, with the development of technologies such as 5G, while ensuring wide area signal coverage, transmission speed and stability have also been improved. The guarantee of information transmission has laid a very solid foundation for the information management of production systems and even all business-related management systems in the future.
These upgrades and changes have actually significantly reduced the barriers to the adoption of autonomous driving in the industry.
We can expect that with the increasing adoption of autonomous driving technology in the industry, there may be significant changes in the future mining process standards, construction management methods, and even design requirements.
InvestMETS: You mention 5G. In terms of enabling technologies, what have been the building blocks to enable you to advance so quickly in the auto-mining field in China?
Lei Zhang: When it comes to rapid development, it is necessary to mention the stimulating effect of various autonomous driving competitions that began in 2016 on the development of China’s autonomous driving industry.
With the popularity of autonomous driving in China, software and hardware has developed rapidly.
On the hardware level, sensors such as LiDAR, automotive-grade high computing power chips, and domain controllers not only continuously improve their capabilities, but also continuously reduce their costs.
At the software level, planning and control algorithms, visual recognition algorithms, simulation tools, etc, are all continuously iterating. Simply put, the basic technological capabilities and industrial maturity of new artificial intelligence provide fundamental support for the rapid development of the field of autonomous driving.
In addition, in terms of vehicles, China’s new energy industry is at a leading level in the world.
With the development of related technologies at large-scale level – that is, suitable for large vehicles – the foundation of autonomous driving vehicles, such as drive-by-wire controlled chassis, electronic and electrical architecture, driving systems, braking systems, etc, has already established a good industrial foundation.
The vehicle can effectively implement functions and technologies in autonomous driving.
Finally, there are developments related to communication.
Mining autonomous driving requires not only single-truck onboard intelligence but also fleet management and dispatching. The development of communication technologies such as 5G, V2V, and V2X in China ensures better management and dispatching through multiple communication channels. This translates into a capability to achieve coordinated scheduling of vehicle-to-vehicle and vehicle-to-systems within the wide coverage of mines.
Taking EACON as an example, while utilising the 5G network to connect devices within the cluster, our auxiliary equipment and trucks are also conducting V2V communication, providing redundant links and ensuring safety while also providing better technical support for the autonomous dispatching and planning.
InvestMETS: Where do you see the most exciting or interesting developments in the sensor/IoT tech field, and why?
Lei Zhang: Mining is a non-standardised scenario, with no obvious lane markings or landmarks, and the physical world changes almost constantly as operations progress.
The improvement of sensor accuracy can help the autonomous driving system more accurately identify the characteristics of obstacles and road boundaries for a better restoration of the physical world, enabling artificial intelligence to have a clearer semantic understanding.
It can be said that the development of sensor hardware performance and processing algorithms will bring significant improvements to autonomous driving.
The development of IoT technology is also very important to autonomous driving.
While improving the system robustness, it greatly reduces the system latency. Thanks to the development of the IoT and edge computing, autonomous driving does not have to focus all decisions on the server side, and the vehicle side has a strong information processing ability.
IoT technologies such as V2V and V2I can achieve direct communication, reduce the probability of data loss, reduce the latency and reduce the bandwidth occupation of the whole system. Combined with distributed intelligence, it can improve the adaptability and safety of the entire autonomous driving system.
InvestMETS: Generally, what is your view of the outlook for the domestic auto-mining market this year? What is market sentiment like currently?
Lei Zhang: I think the market outlook for this year is very positive.
From the overall perspective of China’s mining industry, the development potential is very high.
I think the situation is positive not only this year but also in terms of capacity and the size of the economy in the next few years. The number and proportion of openpit mines are also continuously increasing, including coal mines.
“For EACON 2024 is the first year of large-scale landing”
Taking into account the environmental requirements from the policy side, as well as the fact that a large number of mines are located in remote mountainous areas or at higher altitude, the demand for new energy autonomous driving mining trucks will continue to increase throughout the industry.
For EACON 2024 is the first year of large-scale landing, and the market increment will be positive compared to last year. This will be attributed to our tech maturity.
As I mentioned earlier, EACON has set several benchmark cases in the industry, such as coal mines with hybrid electric trucks and quarries with pure electric vehicles, which have gained a lot of recognition from the industry.
Our high tech maturity and increasing market demand, coupled with the market recognition of EACON, give us confidence in 2024.
InvestMETS: Finally, what, if any, technology transfer from other industries, non-mining, is likely to accelerate the acceptance and adoption of vehicle autonomy, or system autonomy, in mining?
Lei Zhang: From a technical perspective, the introduction of new technologies will bring us new inspiration and opportunities. However, these technologies still require a long period of precipitation and accumulation within mining, making it difficult to directly affect the industry in a short period of time.
For example, Robo-taxi’s technical architecture is very advanced, and its perception and control algorithm is highly applicable in its original scenario. However, when it comes to mining scenarios, it still needs to adapt to new requirements, accumulate technical maturity, and regain customer trust and recognition.
Perhaps it is possible to quickly develop autonomous driving products suitable for mining scenarios within 1-2 years. But mining is a whole production and operation system, so newcomers need to learn how to work together with other auxiliary equipment to continuously improve efficiency.
That requires a long time.
EACON has reached a certain stage of development and accumulated a large amount of capability in different mining scenarios. We are confident that in the future we will be the leading autonomy provider for mining applications globally.
