Capital pours into autonomous mining vehicles
Capital is pouring into the autonomous mining vehicle development field. In just the past August, five companies – Yikong IMa, HuiTuo, TAGE, Mengshi, and Borui – announced funding rounds. Yikong IMa raised several tens of millions of US dollars while HuiTuo and Borui both raised CNY200mn each. The latest round of funding for TAGE exceeded CNY100mn, and the total amount raised in its series B funding was CNY400mn.
The increasing interest in the mining field is due to the relative uncertainty and lack of commercial success in the larger autonomous driving market. Mining has higher levels of certainty, lower risks, and is a more attractive field for investors. As attention shifts to the mining field, talent is expected to increase in the coming years.
Yikong IMa is the only startup company focusing on becoming an operator in this field. Its focus on building a heavier asset model has unique considerations for commercializing autonomous mining vehicles. In June, 9Chapter Autonomous Driving published an article explaining their approach to traditional transportation operations, and in August, they interviewed CEO Watson after their latest funding round.
The topics discussed in this conversation included:
- The benefits of a heavy asset model (1): deeper integration with the scenario party
- The benefits of a heavy asset model (2): easier to gain support from partners
- Battery financing leases are the best solution
- How to manage vehicle purchase funds: bond financing
- Future outlook: non-coal scenarios and overseas markets
The benefits of a heavy asset model (1): deeper integration with the scenario party
From the perspective of mining companies, investing a lot of money in a technology solution from an autonomous driving company that is not yet sufficiently mature involves a lot of risk. Therefore, they are very cautious. However, if the autonomous driving company can provide transportation services directly, they will take on the risk themselves, which increases acceptance from the mining company.
For autonomous mining companies, the service-based business model is easier to accept for mining clients in the initial stage compared to selling technology solutions. This enables them to quickly enter the mine zone, deploy testing vehicles, and have a clear advantage in data collection.
Additionally, as the operator of the entire autonomous driving transportation process, the company masters the “scenario knowledge” that comes mostly from practical experience rather than just research. Consequently, being an operator provides better conditions for a more comprehensive and extensive understanding of scenarios compared to just selling a solution, and thus it is easier to develop and improve technical solutions that better meet the needs of scenarios (including automatic driving algorithms and scheduling algorithms).In mining areas, transportation serves the stripping and mining process, so seamless integration with mining activities is essential. This means that self-driving technology companies not only need to understand self-driving technology, but also need to understand stripping and mining as well as the interaction between different equipment. For operators, since they have more opportunities to come into contact with mining activities, there is naturally no “integration problem” as mentioned before.
Technology vendors mainly deliver projects to the owners and need to meet their delivery standards (often just to make a demo). If the customer’s requirements are wrong, they must follow the wrong path. In contrast, operators have clearer goals (in order to do the job better) and also have sufficient authority to define their own requirements.
On the other hand, the safety staff that technology vendors use in testing may not be their own, but arranged by the mine, and may even be part-time staff. This means that safety staff cannot guarantee an immediate response, which will affect testing efficiency. Conversely, operators use their own safety staff, who are their own employees and can be strictly managed, resulting in higher levels of cooperation.
In addition, the vehicles used by technology vendors in testing are mostly provided by the mine, which calls for tested vehicles according to its own production needs. In contrast, the vehicles used by operators are their own, and they can carry out research and development according to their own plans without worrying about their plans being disrupted.
In the self-driving companies in mining, Yikong IMa was established relatively late, but surpassed others in fleet size and driving mileage. Apart from having a strong corporate culture, this is largely due to the above reasons.
Long-term benefits of being an operator include: only a few major scenarios need to be bound to achieve considerable revenue, so there is no need to constantly develop new customers, which is conducive to concentrating resources to hone technology.
Of course, no matter how amazing the technology is, self-driving vehicles are difficult to escape from the complexity of scenarios. In response to this situation, some self-driving companies have proposed an idea of simplifying the scenario to make it more suitable for large-scale operation of self-driving vehicles. However, this idea is not suitable for all companies.
For example, technology solution providers have little say in whether or not to modify the scenario. They can make suggestions, but ultimately it is up to the scenario owners to implement them.
In contrast, operators may be professional transportation subcontractors for engineering companies. Although they do not have ownership of the scenario, they have a certain degree of “say” in how to complete earthwork stripping and transportation tasks after being subcontracted for a business section, while ensuring safety. Therefore, they can collaborate with engineering companies to modify scenarios.
However, regardless of how the scenario is modified, there will still be some extreme scenarios that are not suitable for self-driving transportation. Therefore, traditional manned transportation is still necessary.Currently, YesAuto AI Driving has carried out traditional transportation in a certain mine in Inner Mongolia. From the current perspective, the main value of doing traditional business lies in understanding the scene, accumulating business resources, and reserving talents. However, in the long term, traditional business will also be a necessary supplement to its unmanned driving transportation business. This is similar to the “mixed dispatching” of companies such as Uber and Didi doing robotaxis.
Of course, mixed dispatching is also a “benefit” that only companies that “take over part of the scene” as operators can enjoy.
Benefits of heavy asset mode (II): easier to obtain partner support
As an operator itself, YesAuto AI Driving has a deeper understanding of the scene, so it also has a deeper understanding of what kind of wire-controlled chassis and sensors are needed for mine scenes. The wire-controlled chassis suppliers and LiDAR suppliers have both reflected that YesAuto AI Driving has given the most feedback among their customers in the mine race track. Correspondingly, these suppliers have also given the greatest support to YesAuto AI Driving.
In addition, companies that act as operators also have obvious advantages in obtaining cooperation and support from vehicle maintenance service providers.
Because maintenance providers usually follow vehicle purchasers, and in the case where unmanned driving companies only act as technology suppliers, the vehicle purchasers are mining companies; in the case where unmanned driving companies act as operators, the vehicle purchasers are themselves.
If a technology solution supplier finds that the vehicle has a problem during testing, it is difficult to directly contact the vehicle maintenance provider for processing. Although it is possible to find the maintenance provider through the mining enterprise, the communication cost is not low. On the contrary, if the operator encounters a problem during testing, the maintenance provider can be contacted directly for support, and the response is faster.
According to Watson, the cooperation mode planned with maintenance service providers in the future is not “how much money to pay for one repair”, but “how much money to split for pulling a certain amount of soil”. What is the difference between the two?
“How much money to pay for one repair” is the most common cooperation mode, but in this mode, the maintenance service provider actually does not have enough motivation to do the vehicle maintenance work particularly well-the better the repair is, the fewer the total number of repairs needed, which also means that the income of the maintenance provider is reduced. For the maintenance provider, the best is to control the amount of maintenance at a “reasonable level”-neither angering the customer nor guaranteeing their income.
From the perspective of the operator, the total number of repairs should of course be as low as possible. Because more repairs not only mean an increase in maintenance costs, but also a reduction in vehicle attendance rate and an increase in work losses. Therefore, operators urgently need to explore a mechanism that can reduce the number of repairs.
“Splitting money for pulling a certain amount of soil” is such a mechanism. Under this mechanism, “minimizing the number of repairs” not only meets the interests of the operator, but also the interests of the maintenance provider. The agreement in interests between the two parties enables the maintenance provider to have more motivation to solve problems in advance and improve the vehicle attendance rate.It is reported that the maintenance and repair company cooperating with Tongli in the coal mine of Pakistan Tal calculates according to the transportation volume, with 90% attendance rate and excellent results after three years. Some mining enterprises have adopted this cooperation mode with tire manufacturers and conveyor belt manufacturers. The latter has the power to make better quality of tires and conveyor belts, with fewer maintenance times.
Battery, financial leasing is the best solution
After the large-scale landing of unmanned driving, fuel costs will become the most important cost. The solution to fuel will be highly related to the current trend of electrification.
Currently, the vehicles used by mining unmanned driving companies are mostly fuel vehicles, but in the context of the countdown to the ban on the sale of fuel vehicles, it is certain that fuel engineering machinery will be replaced by electrification.
2020 is considered the first year of electrification of engineering machinery. Under the guidance of the national strategic direction of “Carbon Peak in 2030 and Carbon Neutrality in 2060”, the process of engineering machinery equipment will be accelerated. Recently, the electric loading machines of Borui Dun sold out, which is the embodiment of the trend of electrification in this micro field. In fact, the electrification of wide-body mining vehicles has begun.
According to insiders, in 2021, Tongli’s production of electric wide-body vehicles is expected to be around 100.
Borui Dun already has dozens of electric wide-body vehicles, and Borui Dun also invested in the mining unmanned driving alliance knowledge technology.
Hongwei New Energy, Hongda Times, and others also have sample vehicles of electric wide-body cars.
The above are all mainframe factories, not unmanned driving companies. So, how do unmanned driving companies view the issue of electrification?
With the development of battery technology, the operating cost of electric vehicles is indeed lower than that of fuel vehicles in the future. However, unmanned driving companies that only sell technology solutions do not need to consider fuel costs during operation. In fact, technical solution providers are unlikely to become the leaders of mining machinery electrification, as the customer can choose between fuel and electric vehicles. However, for companies operating like Yikong IMa, it is different. They need to control fuel costs themselves. Watson said that Yikong has plans to purchase electric wide-body vehicles for testing. “For now, electric vehicles still have great advantages in heavy downhill scenes.”
Another force participating in the market of mining engineering machinery electrification is the power battery manufacturer Ningde Times. In 2020 alone, Ningde Times had the following two investments and cooperation related to the mining scene:
At the end of June, it invested 30 million yuan in the mining unmanned driving company Jump Intelligence and established a joint venture company, Jump Times New Energy Technology, with Jump Intelligence.
In mid-December, it jointly established Hongda Times with Hongda Blasting Engineering Group and other companies, positioning itself as a provider of electrification solutions for the mining industry, providing electrification projects for heavy trucks, engineering machinery, energy storage, and other businesses serving mining customers.The goal of CATL is not only to sell batteries to mining scenarios, but to provide battery financing leasing services, which is crucial to its mission. In September 2020, CATL established a joint venture with logistics giant Prologis, named Ningpu CATL, with a total investment of 20 billion RMB. In 2021, Ningpu CATL established joint ventures with other partners, such as Inner Mongolia Guangpu CATL and Huapu CATL. One of the main goals of these joint ventures is to create a “battery bank” that covers multiple scenarios, offering battery financing leasing and battery secondary use services.
Compared to selling batteries directly, battery financing leasing is a more cost-effective cooperation method for both CATL and mining operators. CATL can borrow from banks at a low interest rate and earn a higher interest rate through financing leasing, resulting in a significant interest rate differential. For mining operators, choosing battery financing leasing means they do not need to purchase batteries (which can account for more than 30% of the total cost of the vehicle), significantly reducing financial pressure.
As the only unmanned driving company focused on being an operator in the mining field, iDriverPlus will inevitably use pure electric wide-body vehicles in future commercial applications and require a large number of batteries. Cooperating with CATL’s battery leasing project may be the best solution, as leasing not only reduces financial pressure but also ensures the best battery usage, thereby maintaining operational efficiency.
Now that the battery issue has been resolved, how can iDriverPlus secure the funding for vehicle purchase? Typically, unmanned driving service operators using heavy assets (primarily start-ups) answer this question with “financing leasing.” However, iDriverPlus has a different answer. Watson did not completely rule out the possibility of financing leasing but values cooperation with banks to do bond financing.
Financing leasing provides a line of credit of up to only 400 million RMB for a single company, and usually, only tens of millions of yuan of funding can be obtained through leasing. On the other hand, obtaining a bank loan is easy as long as the creditworthiness is good, and billions of yuan can be borrowed. In addition, the annualized interest rate of financing leasing is mostly around 10%, with the lowest being more than 6%, while bank loan interest rates are 4%-5%.
iDriverPlus is now the leading company in the mining unmanned driving field, and many investment institutions are bullish on the company’s prospects. However, Watson believes that for companies with good prospects and rapidly rising valuations, equity financing is not very cost-effective. If the valuation doubles within a year, the funding cost of the last round of equity financing is actually 100%. By comparison, the cost of borrowing from a bank is much lower.
Watson said that once a bank recognizes a company’s business model, the level of support it provides is significant. This is very important for the company to scale up its fleet.Of course, when working with banks to do debt financing, it is impossible to borrow a lot of money all at once. It takes time to build up credit. Start with a few million, gradually establish a trust relationship, and then move up to tens of millions, hundreds of millions, and even billions.
However, debt financing has a high threshold, so not everyone can do it. The core reason is that banks have strict approval processes before lending. The money must be used for purposes that comply with government regulations, and the company must also have a stable cash flow.
A leading commercial bank visited Yi Kong Zhi Jia’s test base three times before deciding to extend credit to them. They also conducted in-depth research with many project team members to “make sure that you are really doing unmanned driving operations and that the company can achieve a stable cash flow before agreeing to extend credit”.
In addition, banks often require the actual controllers of borrowing companies to provide unlimited liability guarantees, which are highly tied to their personal assets. In this regard, continuous successful entrepreneurs have a natural advantage.
For these reasons, it is difficult for companies with insufficient strength to obtain debt financing. This explains why many unmanned driving companies that follow the heavy asset model choose the more expensive financing lease model.
Watson believes that because bank audits are very strict, obtaining a loan from a leading commercial bank can also be regarded as the bank’s endorsement of the company. This way, negotiations with other small and medium-sized banks in the future will be much smoother.
Watson also mentioned a strategy–introducing a bank as Yi Kong’s strategic investor. This is a win-win strategy: on the one hand, if the bank invests in Yi Kong early, participates in due diligence, and communicates regularly, it will be familiar with the company and will massively reduce the cost of trust when negotiating loans in the future; on the other hand, the bank can also lock in a potential large customer in advance.
Usually, the purpose of banks’ equity investments is not just to benefit from the increase in the target company’s valuation, but to lock loan customers, which is its core purpose. However, the two are also related. For example, if the target company’s valuation rises rapidly, the bank’s return rate on equity investment will be higher, which will make its approval of loan amounts more active.
This is not just a simple idea. In fact, there are precedents for banks investing in unmanned driving companies. For example, Zhongyin International invested in winning together.
I have had exchanges with some friends who have made investments in banks, and they all believe that under the same circumstances, bank capital has a higher preference for heavy assets.
The reason is that, on the one hand, banks believe that companies with heavy assets are more likely to establish barriers through spending money. On the other hand, banks’ thinking is that the risks of heavy asset companies are “more controllable”. In the bank’s view, it is too uncertain for a technology company to take out a loan to do software algorithm research and development, but if it is used to buy equipment, even if the company dies, the equipment still exists.Previously, I had a conversation with Xiao Yiting, a managing partner at Chentao Capital, about why most self-driving start-ups in specific scenarios prefer a light-asset model, whilst Yuan, the founder of E-Kuaizhi, insists on a heavy-asset model. Xiao Yiting stated that “I firmly believe heavy assets are the barriers. The perception that heavy assets are “risks” mainly stems from a lack of control over finance and cash flows, or an inadequate ability to assess risks. However, I have experienced building from scratch to going public and have over 10 years of experience in operating heavy assets.” This viewpoint is also applicable to E-Kuaizhi.
Looking to the future, E-Kuaizhi plans to expand its scenes to non-coal mining settings and foreign markets, as the leading investor in its latest financing round is the global non-ferrous metal mining giant, Zijin Mining. E-Kuaizhi stated: “E-Kuaizhi is expected to expand its scenes to non-coal mining in the near future, such as non-ferrous metal mines…Zijin Mining is also expected to empower E-Kuaizhi in multiple aspects, such as non-ferrous metal mine scenes, EPC cooperation partner in mining, and going international.” The crucial words here are “non-coal scenes” and “going international.”
Currently, the market volume of transportation of overburden in China’s open-pit coal mines is 25 billion yuan per year, that of sand and aggregates mines is 56 billion yuan per year, that of open-pit non-ferrous metal mines is 21 billion yuan per year, and that of open-pit iron mines is 28 billion yuan per year. If the overseas market is taken into account, the total amount is not insignificant.
So, when is the best time for E-Kuaizhi to expand into non-coal mining settings? The answer, according to Watson, is 2023. According to the plan, E-Kuaizhi will gradually remove safety officers starting from the end of 2021 and then enhance safety and reliability. By 2023, its technology for open-pit coal mines will be relatively mature, and the fleet size in a single large mine will reach about 80 vehicles. The focus will then shift from autonomous driving algorithms to scheduling optimization algorithms to improve operational efficiency. Therefore, this is a suitable time to delve into non-coal mining settings.
Zijin Mining will introduce EPC partners it has been cooperating with for years to E-Kuaizhi, and the two parties will form a strategic alliance.
In the medium and long term, cooperation with Zijin Mining will also reduce the difficulty of E-Kuaizhi in expanding into overseas markets. E-Kuaizhi already has plans to expand into overseas markets after its success in the domestic market, and Zijin Mining’s high-quality mines overseas are likely to become E-Kuaizhi’s first stop on the road to going international.
Australia will be E-Kuaizhi’s first overseas market. The reasons are as follows:
-
Australia has mining giants such as Rio Tinto and BHP Billiton, and these giants have strict management of mining operations. If E-Kuaizhi can cooperate with these giants and build one or two benchmark projects of unmanned driving operations, it will be much easier to enter other mines around the world in the future.2. The cost of Australian drivers is very high, with dump truck drivers earning over 200,000 Australian dollars annually. They usually commute by helicopter and live in three-star hotel-standard dormitories. With paid annual leave, the annual cost of a driver is around 2 million yuan. Therefore, Australian mining companies are more motivated to invest in autonomous driving technology.
-
Although Australia has many good universities, there are few technology companies, resulting in an excess of talent supply. Therefore, algorithm and engineering talents are relatively easy to recruit and cost similar to China.
During the communication with Caterpillar experts, Eicontrol learned that in fact, the technology of domestic unmanned mining companies is very close to or even not inferior to Caterpillar and other mining truck giants. Therefore, Watson believes that they have a great opportunity to compete with Caterpillar and Komatsu in the international market. Especially, the Chinese are more hardworking and have an advantage in operating mines, which require dirty and tiring work.
Compared with Caterpillar and Komatsu, Eicontrol Intelligence has a greater advantage in cost when developing the Australian market.
At present, Caterpillar and Komatsu use large mining trucks with a carrying capacity of more than 200 tons and a price of around 20-30 million yuan, while Eicontrol Intelligence uses wide-body trucks with a carrying capacity of 40 tons and a price between 600,000 and 900,000 yuan. As a comparison, the carrying capacity of the wide-body truck is about 1/6 of the mining truck, and the price is about 1/40 of the mining truck.
At present, due to the relatively high cost of drivers in Australia and other countries, large mining trucks can greatly reduce the demand for drivers. Therefore, overall, large mining trucks still have certain advantages. However, when everyone implements unmanned driving and does not require drivers, the advantage of “saving drivers” of large mining trucks will no longer exist, and the cost advantage of wide-body trucks will be highlighted.
Currently, Tongli’s wide-body trucks have already been exported, and in the future, they will be sent to the Australian market with Eicontrol’s autonomous driving solution, which is not very difficult.
However, it is difficult for wide-body trucks to completely replace large mining trucks, and even if they only partially replace, it will not happen quickly. This is because the operating methods of existing mining areas in Australia are designed based on the characteristics of large mining trucks. Switching to wide-body trucks means that their operating methods also need to be modified, which is almost impossible. Therefore, in the initial stage, Eicontrol’s plan based on wide-body trucks can only target newly opened mines and focus on incremental markets.
Of course, if wide-body trucks can set a new benchmark in new mines, there is still a great chance to enter old mines for modification. After all, the bidding for mining areas also needs to consider overall economic benefits. If the cost saved by replacing large mining trucks with wide-body trucks is higher than the cost of technological transformation, why wouldn’t mining companies do it? And large mining trucks also have a scrap cycle, and can be gradually replaced with wide-body trucks after the used stock of mining trucks is discarded, without much conflicting interests.In 2023, Easy Control plans to conduct unmanned driving tests and trial operation in Chinese-owned mines in Australia, and in 2025, it plans to cooperate with giants such as Rio Tinto and BHP Billiton to engage in large-scale commercial operations. Subsequently, it will expand into the Canadian market.
In addition to the Australian and Canadian markets, Easy Control also plans to cooperate with Tongli to develop the African and Asian markets.
At first, when the author heard about this plan, he didn’t quite understand: “The cost of drivers in African and Asian countries is very low. Unmanned driving may be more expensive, so does this market really exist?” Watson explained that unmanned driving is not replacing local drivers but the drivers sent from China.
In fact, domestic unmanned driving companies mainly serve Chinese mining enterprises with industries in Africa and Asia, but the drivers used by Chinese companies in local mines are not the “cheaper” local drivers, but the ones hired from China. Local drivers may be cheaper, but they are far less diligent than Chinese drivers and work inefficiently, which seriously affects the attendance rate of vehicles.
Moreover, if the driving habits of the drivers are not good, it can easily cause damage to vehicles. Therefore, some mining companies that have done business in African countries have found that five local drivers are not as good as one Chinese driver. Therefore, even if Chinese drivers are much more expensive, they are more inclined to use them.
However, recruiting drivers from China and then sending them to African and Asian countries usually requires a monthly salary of 20,000 yuan, plus round-trip tickets and annual leave. The cost of a single driver can be as high as 300,000 yuan per year. If there are many drivers, this will be a huge expense. Therefore, Chinese companies mining in African and Asian countries are also highly motivated to cooperate with unmanned driving companies.
This article is a translation by ChatGPT of a Chinese report from 42HOW. If you have any questions about it, please email bd@42how.com.