2018年10月4日，美国交通运输部发布了备受期待的《自动驾驶汽车3.0：准备迎接未来交通》，也就是业内为人所熟知的AV 3.0指南。

该指南在美国交通运输部 2017 年出台的《自动驾驶系统2.0：安全愿景》（AutomatedDriving Systems 2.0: A Vision for Safety）基础上，纳入自动驾驶行业的各方意见，包括汽车制造商、供应商、各州和地方政府以及基础设施所有者和运营商等。

这些指南产生了一个概念上的矛盾，因为它们既提醒各州了解各自的权力有限，却又对汽车制造商和供应商不加约束。

看起来，AV3.0 指南具备一定限制含义，但又同时留有余地。对于美国交通运输部新发布的这一指南，汽车行业将如何应对？

不是强制规则，而是指导方针

与联邦机动车辆安全标准（FMVSS）不同，AV3.0指南本身的性质为指南，而非标准或规定。迄今为止，美国国家公路交通安全管理局（NHTSA）发布的所有自动驾驶汽车相关指南都不具备强制性，旨在鼓励各州和地方政府及私营企业自愿采用这套规则，促进自主技术的健康发展。

NHTSA（隶属于美国交通部）、FMVSS 标准及其他相关汽车监管框架的起源均可追溯至 1966 年“国家交通和机动车辆安全法案”（theNational Traffic and Motor Vehicle Safety Act of 1966）。制定之初，这套安全法案的目标是通过对安全要素的监管，“减少交通事故数量”。当时，这种“由NHTSA创建安全汽车框架”的规定制定策略甚至被誉为“现代政府最伟大的发明”。当时，制定汽车行业规定主要有三个关键目标：（1）驱使行业保持创新；（2）确保驾驶员和乘客的基本驾乘安全；（3）将安全的重要性置于成本之上。

当时，除了颁布行业规定之外，NHTSA还被赋予了“监督车辆召回”的合法权利。

自此之后，NHTSA立即开始工作，并陆续发布了很多我们今天所熟知的安全标准。然而，在1974年到1986年的十多年中，NHTSA并未发布任何重大FMVSS标准。事实上，NHTSA在最初推行新规时受到了行业各方的大力反抗，新规也受到了“不合理、不实际、不客观及未以安全为导向”等一系列指责。

从结果来看，第一回合的“胜利者”是企业。在成立的最初10 年中，NHTSA共制定了 10 项主要规定，但6 项均以失败告终，其中包括有关“被动约束系统”和“防抱死制动系统”的规定。与此同时，NHTSA在监管车辆召回方面，反而取得了更为有利的“战果”。

在出台主要规定接连失利之后，NHTSA再未开展重大行动，直到美国国会提出强制要求。1991年，美国国会通过了ISTEA法案，要求有关车辆防侧翻、侧面碰撞保护、助推器座椅设计改进及一系列相关标准就位。2000年，比尔·克林顿总统签署了《有关加强运输车辆召回、问责和文件记录法案”》（Transportation Recall Enhancement,Accountability, and Documentation Act，即TREAD法案），主要关注轮胎问题。历史经验表明，如果国会不出手，NHTSA可能更倾向于对FMVSS标准及相关规定保持沉默。事实上，从某种程度而言，AV 1.0、AV2.0及AV 3.0的情况都是如此。

不过，AV3.0指南的情况略有不同，最起码表明了NHTSA多年以后重新进入规则制定领域的意向。

NHTSA 是否会试图消除“额外监管障碍？”

最新AV 3.0 指南已经认识到，现行的FMVSS标准可能成为阻碍创新的障碍。

专为SAE4级和5级自动驾驶设计的自动驾驶汽车，可能不会配备方向盘、刹车、油门、后视镜等零部件，而这明显不符合现行FMVSS 标准对道路车辆的要求。在未来制定新规的过程中，NHTSA开始就“修改相关安全标准”征求意见。2015年11月12日，谷歌曾致函要求对FMVSS标准中有关“车辆必须配备驾驶员”的规定进行解释，并表示“配备自动驾驶系统”的无人驾驶汽车也应被视为“已配备驾驶员”。根据当时的回复，我们不难推断，NHTSA未来将重新审阅大量FMVSS标准。

除了新规制定流程，NHTSA还希望通过一系列努力让FMVSS标准更加灵活、响应迅速、以性能为导向，但同时不区分具体技术。AV 3.0指南表明，未来的交通环境中可能充满各种不可预测的障碍物、车辆和行人，在此环境下，新时代的FMVSS在“验证自动驾驶系统时可能采用一套更简单和更通用的标准”。

无论如何，这些建议都预示着一个信息：未来，NTHSA将不再继续沉默。

如何防止各州安全标准的“东拼西凑”？

值得一提的是，AV3.0指南再次强调了FMVSS标准的优先权。简单来说，当联邦法律“压制”或优先于与其不一致的州法律时，就以联邦法律为优先。2000年，美国最高法院曾在一起诉讼中分析了这一问题：该诉讼称一款1987款本田的被动约束系统未配备安全气囊，因此应视为设计缺陷；而当时的FMVSS标准允许制造商在设计被动约束系统时三选其一，其中包括安全气囊。在此案例及随后的几个相似案例中，法院均裁决州、地方和部落政府不能通过立法对车辆的安全性能做出与联邦标准不同的要求。

事实上，在所有有关自动驾驶汽车的指南方针中，NHTSA早已明确表示其负责制定安全和性能相关标准，而各州继续负责其它四个关键领域：保险、执照、交通安全法和责任（包括产品责任）。

未来路在何方？

AV 3.0指南中唯一可以确定的部分是：自动驾驶车辆开发和部署制度仍处于自愿遵守阶段。从AV 3.0指南附录中纪录的大量研究中，我们可以清楚看到，NHTSA将继续高度依赖SAE等标准组织所做的工作，并同时回馈给他们可以激发行业创新的规定。我们相信，正如TREAD法案曾推动解决轮胎相关缺陷问题一样，立法也将加快NHTSA推行自动驾驶汽车规定。

然而，如果目前仍在分裂之中的国会未能及时采取行动，NHTSA可能将不得不继续在新规制定过程中不断试探。不过，曾经的问题仍然存在：汽车行业是否会支持NHTSA的努力？是否又会像1968年至1974年间那样，疯狂质疑任何新起草规则的权威和适用范围。这次，考虑到不确定性的存在和豁免权的限制，行业可能会做出不同选择！

On October 4, 2018, the U.S. Department of Transportation released the long-awaited “Autonomous Vehicles 3.0: Preparing for the Future of Transportation” guidelines—otherwise known as AV 3.0.

These guidelines build upon the voluntary guidance issued in the DOT’s 2017 “Automated Driving Systems 2.0: A Vision for Safety” and incorporate comments from autonomy stakeholders that include auto manufacturers, suppliers, state and local governments and infrastructure owners and operators.

These federal guidelines create a conceptual challenge, as they remind states of the limits of their power while being—controversially—non-binding on automotive manufacturers and suppliers.

Potentially restrictive, yet simultaneously optional. How can the industry navigate this recent DOT release?

It’s not a rule; it’s a guideline 

Unlike the Federal Motor Vehicle Safety Standards (FMVSS), the AV 3.0 guidelines are just that: guidelines. All of the AV-related guidelines issued by the National Highway Traffic Safety Administration to date are voluntary and serve to encourage state governments, local governments and private industry to adopt the principles to promote the development of autonomous technology.

The origins of NHTSA (part of the DOT), the FMVSS and the related automotive regulatory framework are found in the National Traffic and Motor Vehicle Safety Act of 1966. The goal of the Safety Act was to “reduce traffic accidents” through the regulation of safety elements. This rulemaking strategy, where NHTSA would create the framework for safer vehicles, was even hailed as one of the “greatest inventions of modern government.” This approach to auto-industry regulation would lead to three key objectives: (1) compel the industry to strive for innovation; (2) ensure that drivers and passengers would be reasonably safe in their vehicles and (3) elevate safety above cost.

Along with this rulemaking authority, NHTSA was granted the legal rights to oversee recalls.

The agency immediately went to work and issued many of the safety standards we have come to know today. Yet in the decade-plus from 1974 to 1986, there were no major FMVSS issued. After the initial push, companies pushed back, claiming that many of the agency’s rules were not reasonable, practical, objective, or safety-oriented.

Industry won the first battles. During the first 10 years of its existence, NHTSA lost six of its ten major rulemaking cases including cases addressing passive restraints and antilock brakes. At the same time, NHTSA received far more favorable outcomes from cases addressing its recall-enforcement mandate.

Following these significant rulemaking losses, NHTSA acted in only minor ways until compelled by Congress. In 1991, Congress passed ISTEA, a bill requiring rulemaking address rollover prevention, side impact protection, booster-seat design improvements and related standards. In 2000, President Bill Clinton signed the Transportation Recall Enhancement, Accountability, and Documentation Act (TREAD) which focused predominantly on tire issues. Absent Congressional action, history suggests that NHTSA will tend to remain silent on FMVSS and related regulation. This was true of AV 1.0 and AV 2.0 and, to a degree, AV 3.0.

The only difference is that AV 3.0 suggests NHTSA has its eye on reentering the rule making arena after these many years.

Will NHTSA seek to remove an “unintended regulatory barrier?” 

The new AV 3.0 guidelines recognize that the current FMVSS may be a barrier to innovation.

Autonomous vehicles designed for SAE Level 4 and 5 performance may not be equipped with steering wheels, control pedals, mirrors or other components that are currently required for on-road vehicles. The NHTSA will, in an upcoming rule making, seek comment on changes to relevant safety standards; judging by the agency’s response to Google’s November 12, 2015 letter requesting interpretation of the FMVSS as applied it a driverless vehicle operated by a “Self- Driving System,” it is likely a review of any number of existing FMVSS.

In addition to new rule making, NHTSA will make FMVSS more flexible, responsive, technology-neutral and performance-oriented. AV 3.0 suggests that new-age FMVSS may “incorporate simpler and more general requirements to validate” autonomous systems in a world where unpredictable obstacles, vehicles and pedestrians dot the landscape.

These recommendations foreshadow a departure from the recent historical silence that has fallen over NTHSA rule making.

How to prevent a patchwork of state safety standards? 

The guidelines continue to address the preemptive force of the FMVSS. Federal preemption occurs when a federal law “trumps” or preempts a state law that is inconsistent with the federal law. In 2000, the United States Supreme Court analyzed the issue of preemption in a lawsuit that asserted that a 1987 Honda had a design defect because it did not have airbags at a time when the FMVSS allowed manufacturers to choose one of three options (including airbags) to address passive restraints. Through this case and several that followed, courts determined that state, local and tribal governments cannot enforce laws on the safety performance of a vehicle in ways that differ from federal standards.

In all of the guidance issued on autonomous vehicles, NHTSA has been clear that it holds responsibility for establishing safety and performance-related standards, while the states retain four key areas: insurance, licensure, traffic safety laws, and liability (including product liability).

What will the future hold?

The only certainty given in AV 3.0 is that the regime for autonomous-vehicle development and deployment is still in the voluntary-compliance phase. It is clear from the significant research placed in AV 3.0’s appendices that NHTSA will continue to rely heavily on the work performed by standards organizations such as SAE International while they hope to return to the rule making function that inspired their creation. Perhaps legislation will speed that effort, just as the TREAD Act did for tire-related defects.

But in the event the current divided Congress fails to act, it is plausible that NHTSA will once again test the rule making waters. The only question that will remain is whether the industry will support those efforts or, as it did from 1968-1974, challenge the authority and the scope of any newly drafted rules. Perhaps with the existing uncertainty and limits in exemptions, the choice may be different this time.

Author: Jennifer Dukarski

Source: SAE Autonomous Vehicle Engineering