Author: Su Qingtao
When the driver completely hands over the control of the vehicle to the onboard computer system of the autonomous driving vehicle, the steering wheel, accelerator, and brake are all electronically controlled (previously through hydraulic-mechanical means), which is called “drive-by-wire” control. Drive-by-wire control mainly includes throttle-by-wire, steer-by-wire, and brake-by-wire.
Because electronic signals are faster than mechanical connections, drive-by-wire can provide higher level safety protection for autonomous driving. For example, the response time of a conventional braking system is 300-500 milliseconds, while the response time of the iBooster is 120-150 milliseconds, and the response time of the Brake-by-Wire system of Brabham is only 90 milliseconds, which shortens the braking distance of drive-by-wire. Continental claims that when the pedestrian protection is activated at 30 km/h, the MK C1 braking distance can be reduced from 6.8 meters to 4.1 meters.
Brake-by-wire is an execution layer component. The generation of braking signals can come from the pedal. After the pedal stroke sensor measures the displacement of the input push rod, it sends the displacement signal to the ECU, and the ECU calculates the braking request. It can also be generated by the ECU according to the scene’s needs.
Among the three independent drive-by-wire systems, throttle-by-wire has the highest popularization rate. In vehicles with ACC and TCS functions, throttle-by-wire has become a “standard configuration”. Brake-by-wire and steer-by-wire have a less favorable consumer experience due to the immaturity of early technology. Also, drive-by-wire technology is difficult to attribute responsibility as it adjusts and controls the execution mechanism through the car’s computer. All these factors have hindered their popularization and promotion in the market.
Among them, brake-by-wire is the most critical and difficult. The rapid development of intelligent connected vehicles in recent years has brought new opportunities for drive-by-wire technology.
Before the era of autonomous driving, suppliers successively launched the following brake-by-wire products for traditional vehicles:
-
The ECB developed by EDEX was applied to the Toyota Prius from 1997 onwards;
-
The SBC developed by Bosch has been applied to Mercedes-Benz CLS, SL, and E-Class since 2001;
-
The SBC developed by Continental was applied to the Fusion and Milan of Ford in 2009;
-
Brembo’s BbW, developed by Brabham, has been applied to many F1 cars since 2014.
Without exception, these products had quality defects. Brembo BbW caused major accidents in F1 races for three consecutive years, and the other products have also caused recalls of tens of thousands or hundreds of thousands of vehicles.Currently, the several wire-controlled braking products that have caused accidents have been phased out. The same-named products of ATEKES and Brembo which are still in use have undergone revisions.
The development of braking products has gone through three generations so far. The first generation was the mechanical braking system, followed by the engine-assisted braking system. The third-generation product is the wire-controlled braking system that is not dependent on engine assistance but rather electric power and digital control.
The fourth-generation braking product will be wire-controlled with redundancy mechanisms, mainly developed for autonomous vehicles.
Currently, Bosch, Continental, ZF (including TRW and Wabco), Hitachi (including Fambro Brake), ATEKES, and Brembo are the major companies that can supply or will soon supply wire-controlled braking products suitable for autonomous driving vehicles.
Table: Domestic and foreign wire-controlled braking product solutions
In terms of technology, compared with the traditional mechanical braking method, the main advantages of wire-controlled braking are as follows: 1. It responds faster and can brake in a shorter time. 2. It has a simpler structure and lighter weight. 3. It has strong energy recovery capabilities and effectively utilizes the energy generated by friction during braking, extending the driving range. 4. It has a backup braking system that provides redundancy.
However, at present, in terms of the actual performance of wire-controlled braking, it is still acceptable for L2 level autonomous driving vehicles, but if used to support L4 level autonomous driving, it will face significant challenges.
Among the several wire-controlled braking products listed in the above table that can be used for autonomous driving, the IBC of ZF (TRW) has just been installed on vehicles and has not been verified on a large scale. The Smart Brake of Hitachi (Fambro Brake) has not yet been mass-produced, while several others that have been mass-produced and installed, such as Bosch’s iBooster, Continental’s MK C1, and Hitachi’s E-ACT, have all experienced accidents of varying degrees over the past few years.
Bosch’s iBooster is currently the wire-controlled braking product with the highest market share. In the product design, iBooster and ESP serve as braking redundancies, which to some extent satisfy the needs of autonomous driving.
However, ESP, as the braking redundancy, is still a traditional electro-hydraulic device, requiring three times the braking time of the main braking system, such as iBooster. Moreover, each use requires the plunger pump to withstand high temperature and pressure. Frequent use will cause severe heating of the plunger pump, reduce precision, and dramatically shorten the lifespan of ESP.But in the accident in the Honda CR-V, when the iBooster failed, the redundant ESP also lit up the fault light – this means that both the main braking system and the redundant braking system failed at the same time! This “double unreliable” not only cannot be applied to L4-level autonomous driving, but even for L2-level autonomous driving, it is somewhat difficult.
L4-level autonomous driving must have electronic redundancy and cannot rely solely on mechanical redundancy, otherwise the driver will be trapped in the quagmire of responsibility for taking over the vehicle in a period of time. To address this potential risk, Bosch has launched the One Box solution based on iBooster – a highly integrated product IPB that integrates the functions of iBooster and ESP, equipped with RBU as a braking redundancy. This realizes the dual safety failure modes of mechanical redundancy + electronic redundancy.
The IPB+RBU solution with dual redundancy support can support L3 and L4 level autonomous driving.
When Continental’s line control brake MK C1 was supplied to Alfa Romeo, the production process issue of the OneBox solution had not been resolved, and thus brake redundancy was not available. If MK C1 fails, the system will remind the driver through a warning light. The driver is the ultimate “braking redundancy” solution and can brake by stepping on the pedal.
By 2017, the production process issue of the OneBox solution was close to resolution, and Continental launched the OneBox solution that can meet the requirements of L3 and above autonomous driving, which integrates MK C1 and the redundant system MK100 HBE into the same box. If the main brake system fails completely, the MK100 HBE unit will use the two front wheels to brake the car, playing the role of an anti-lock brake system.
If the electromechanical actuator and pump functions of the main brake system fail, but the control valve is not affected, the MK 100 HBE unit will enter the collaborative brake mode, with some of the hydraulic pressure sent to the functional valve of the stationary MK C1 to drive the rear brake system.
Continental has specifically mentioned that MK C1 is “suitable for advanced autonomous driving” and “does not require manual intervention during the braking process” on its official website, and at the 2019 Shanghai Auto Show, Continental China CEO Tang En announced that MK C1 can meet the requirements of L4 level autonomous driving (but due to cost reasons, the MK C1+MK100 HBE full kit solution currently has no publicly available production orders).
But the highest level of other line control brakes from other companies can only support L2 at present.The official website of ZF states clearly that its intelligent braking system (IBC) is suitable for “semi-automated driving”. Ajey Mohile, Chief Engineer of ZF-TRW, said in an interview with the media that the IBC has “no real redundancy”.
Valeo is also straightforward: Valeo’s Electric Brake System (EBS) “does not have redundancy and needs to be redundant with other systems such as EPH. Currently, it can support up to L2 at most”.
Aixeed’s official website clearly states that its Electronic Brake Control (EBC) can support ADAS, which means it is not exceeding L2. Considering that Toyota’s focus on autonomous driving is on L2, Aixeed may not have strong motivation to launch a line control system that can support L4 for a considerable time in the future.
In addition to the aforementioned suppliers, companies such as Autoliv, Nissin, WABCO, and Mando can also provide line control products, but they are still far from meeting the needs of autonomous driving.
In November 2020, the State Council of China issued the “New Energy Vehicle Industry Development Plan (2021-2035)”, which explicitly proposed to overcome the line control execution system as the “neck” core technology.
Among Chinese companies, Wanxiang Qianchao, Wanan Technology, Asia-Pacific Mechatronics, Nosen, and Berkeley are all researching line control technology. Among them, Asia-Pacific Mechatronics’ IEHB has been integrated into the (BAIC Yinxiang) vehicle, Nosen’s Nbooster has been successfully installed on the BAIC EC3 in January 2019, and Berkeley’s WCBS is also planned to start mass production in May 2020.
Berkeley has mature and complete ABS, ESP, and EPB technologies. Especially, the EPB technology can serve as the electronic redundancy of the line control, and the One-box solution is ahead of other domestic manufacturers in terms of progress.
These suppliers’ line control products have all set their sights on the autonomous driving market, but the maturity and reliability of the products still need to be verified by the market.
The One-box solution has become the mainstream trend.
The definition of the One-box solution and the Two-box solution lies in whether the AEB/ESP system is integrated with the electric power steering in one place. In the Two-box solution, the redundant ESP and electric power steering are independent of each other, while in the One-box solution, the electronic processor itself integrates ESP.
The One-box solution can achieve higher energy recovery efficiency, and due to its high integration, the volume and weight are greatly reduced, and the cost is also lower. However, there are many technical challenges, such as the need to decouple from the pedal. Because the pedal is only used for input signals and does not act on the master cylinder, the sensor senses the pedal force to drive the motor to push the piston, and the pedal feeling needs software tuning, which may have safety risks.These technical challenges resulted in a delayed production time for the one-box solution. For instance, Bosch’s Ibooster utilized the two-box solution for its first and second generations, and only the latest IPB uses the one-box solution. ZF TRW’s EBB belongs to the two-box solution, while Continental’s MK C1 and Wabco’s WCBS directly adopt the one-box solution.
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.