Self-Driving Delivery In Mine Enters Into Its Era

“Removing safety officer” has always been an important prerequisite for the commercialization of autonomous driving above level L4. For any autonomous driving company positioned above L4, it is an important node and a new starting point for their development. At present, unmanned normal operation is achieved in completely closed road sections, such as mine scenarios.

The motivation of the market

Mining is divided into open pit mining and underground mining. Usually, the decision to open pit mining is made based on factors such as stripping ratio (the amount of waste rock removed per unit of useful mineral extracted), ore quality, and price. At present, almost all domestic autonomous driving enterprises focus on the transportation scenarios of open-pit mines.

On the demand side, there is a strong desire for new technologies and a new mode.

There are two core reasons. The first, of course, is the labor problem basically in all autonomous driving vertical commercial scenarios.

There is a strict requirement for the drivers who are allowed to work in mines: only those with license B or above can drive the off-road wide-body dump trucks (currently used in the mines). In addition, the relatively harsh working environment of the mines located in remote areas and the double-shift work make the driver position less attractive.

Another important point is security.

At present, the largest proportion of safety accidents in open-pit mines occurs in the transportation link. As the mine is an industrial production system, even if the driving accident was caused by the driver’s own mishandling or dozing off while driving, the management cannot shirk the responsibility. At present, it is mainly avoided through digitalization or fine management. For example, drivers who are on the night shift will be forced to stop for rest between 3:00 am and 4:30 am.

“Safety” is the top priority for mines that requires volume target. Once related problems occur, we will face work suspension. If the approved capacity cannot be completed, we will see huge losses.

Therefore, from the perspective of labor cost and safety, mine operators have a strong demand for self-driving delivery technology.

The technical feasibility

Once the external environment is satisfied, we should focus on feasible and sustainable development.

Scenario-wise, people who enter the mine site have received a lot of on-the-job training, and there will be no uncontrollable personnel (such as children who often appear in the terminal logistics scene). This means that the corner case within the whole closed scenario is in a controllable area, it is relatively easy to achieve the dual guarantee of efficiency and security. At the same time, another advantage of strong closure is that some problems that cannot be solved in the short term can be solved by manual management.

Data-Centric

Poor quality data annotations often lead to poor model performance, so the quality of data annotations is critical.

Let’s take a look at an object detection data annotation case in a mining area.

Labeling categories

Cars (including cars and pickup trucks), trucks (including mining cards, watering carts, and tankers), pedestrians ( including pedestrians, riders, and those holding umbrellas), excavators (including stone drill), bulldozers (including bulldozers, forklifts, and rollers) and stones.

Labeling size

Small objects(take stone as an example): with a minimum annotation pixel of 12x12 and the maximum annotation pixel of 150x150, label it as the target is visible to the naked eye, and its category, can be determined.

End

Outsource your data labeling tasks to ByteBridge, you can get high-quality ML training datasets cheaper and faster!

• Free Trial Without Credit Card: you can get your sample result in a fast turnaround, check the output, and give feedback directly to our project manager.
• 100% Human Validated
• Transparent & Standard Pricing: clear pricing is available(labor cost included)

Why not have a try?

Source:https://baijiahao.baidu.com/sid=1743825116872849966&wfr=spider&for=pc