Policy ForumTechnology Development

Autonomous vehicles: No driver…no regulation?

See allHide authors and affiliations

Science  06 Jul 2018:
Vol. 361, Issue 6397, pp. 36-37
DOI: 10.1126/science.aau2715

eLetters is an online forum for ongoing peer review. Submission of eLetters are open to all. eLetters are not edited, proofread, or indexed.  Please read our Terms of Service before submitting your own eLetter.

Compose eLetter

Plain text

  • Plain text
    No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Author Information
First or given name, e.g. 'Peter'.
Your last, or family, name, e.g. 'MacMoody'.
Your email address, e.g. higgs-boson@gmail.com
Your role and/or occupation, e.g. 'Orthopedic Surgeon'.
Your organization or institution (if applicable), e.g. 'Royal Free Hospital'.
Statement of Competing Interests
CAPTCHA

This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.

Image CAPTCHA
Enter the characters shown in the image.

Vertical Tabs

  • International standards are needed for securing autonomous vehicles

    Joan Claybrook et al. wrote an article “Autonomous vehicles: No driver ... no regulation?” (1). Andrew Robinson also wrote an article “Looking ahead” (2). Both articles agree that autonomous vehicles (AVs) will play a crucial role in our society in the self-driving age. However, current AVs have unsolved security vulnerabilities so that we must overcome the vulnerabilities and resolve all the problems. I have collected the known security vulnerabilities for AVs or connected vehicles (CVs) (3). There are two kinds of potential vehicle attacks: forged vehicle communications (in-vehicle network, inter-vehicle network or telematics, and vehicle access attacks), and sensor attacks. We must devise and develop several new technologies for nullifying those potential vehicle attacks. In order to make autonomous vehicles usable in our society, international standards for securing sensors are needed.

    Forged vehicle communications:
    In-vehicle network attacks
    In OBD2 (On Board Diagnostics level 2) standard, no security is embedded or provided. Therefore, in the in-vehicle network communications, security functions including encryption / decryption between ECUs (electronic control units) must be embedded in OBD2 or the OBD2 standard must be replaced with the better standards for AVs or CVs.

    Inter-vehicle network attacks
    In the inter-vehicle network or telematics, since 4G/LTE/3G hijacking has been reported (4,5), the new communications networks including...

    Show More
    Competing Interests: None declared.
  • RE: Autonomous vehicles: No Driver ... no regulation

    The Policy Forum article by Joan Claybrook and Shaun Kildare on Autonomous vehicles begins, as do most articles on this subject, with a discussion of the yearly rate of automobile accident deaths and injuries, and an assertion that some or all of these incidents will be avoided by automation. However, that is not the reason that industry is racing to deploy such vehicles. Industry is pouring billions of dollars, and asking the Congress to short-cut the regulatory path to deployment, because the business case they have developed dictates that large profits will accrue when commercial (human) drivers are replaced by robots. This is important to understand as constituents evaluate the merits of fast-tracking this technology. A reduction of deaths and injuries seems always to be placed first when this is brought up, but make no mistake, the attraction for the industry is profits.

    Competing Interests: None declared.
  • RE: Autonomous vehicles: No driver…no regulation?
    • Dallas Weaver, semi-retired "applied science", Sceintific Hatcheries

    The Policy Forum article by Joan Claybrook and Shaun Kildare on Autonomous vehicles (AV) begins by crediting automation with the promise of saving some of the 37,000 lives per year (with many times that many serious injuries), along with massive economic damage. The authors then go on to cite the failures of the existing AV systems.

    However, the only numbers which really count are the accidents/million miles for the AV vs. human drivers with all their flaws, poor sensors that can be confused, sloppy algorithms and poor mental processing allocation decisions (distractions). Because these numbers are certainly available to the authors -- as part of their expertise, they should know these numbers off the top of their heads -- the question is "Why they weren't presented?"

    If the numbers show that Waymo, etc. are actually safer than the average human driver, the authors' requests for more regulation and funding increases would be questionable. Fewer new regulations slowing down implementation of AV technology would result in more lives saved. Having regulators second guessing AI learning systems could result in FDA-class delays in improvements and the accompanying loss of a huge number of lives. The regulator's image of the "ideal" could kill the "better than human". People will die from their well-meaning perfectionism.

    If the current risk is higher with the AV, the rate of decrease is relevant, and...

    Show More
    Competing Interests: None declared.
  • RE: Autonomous vehicles: No driver .. no regulation?

    The Policy Forum article by Joan Claybrook and Shaun Kildare on Autonomous vehicles (AV) begins by crediting automation with the promise of saving some of the 37,000 lives per year (with many times that many serious injuries), along with massive economic damage. The authors then go on to cite the failures of the existing AV systems.

    However, the only numbers which really count are the accidents/million miles for the AV vs. human drivers with all their flaws, poor sensors that can be confused, sloppy algorithms and poor mental processing allocation decisions (distractions). Because these numbers are certainly available to the authors -- as part of their expertise, they should know these numbers off the top of their heads -- the question is "Why they weren't presented?"

    If the numbers show that Waymo, etc. are actually safer than the average human driver, the authors' requests for more regulation and funding increases would be questionable. Fewer new regulations slowing down implementation of AV technology would result in more lives saved. Having regulators second guessing AI learning systems could result in FDA-class delays in improvements and the accompanying loss of a huge number of lives. The regulator's image of the "ideal" could kill the "better than human". People will die from their well-meaning perfectionism.

    If the current risk is higher with the AV, the rate of decrease is relevan...

    Show More
    Competing Interests: None declared.
  • The security problems of autonomous vehicles are solvable as long as the facts are disclosed

    Joan Claybrook et al. wrote an article entitled “Autonomous vehicles: No driver…no regulation?” (1). Scientists and engineers must disclose the facts of problems on the conventional vehicles and the expected autonomous vehicles. I have recently published a paper entitled “connected vehicle security vulnerabilities” where a variety of security problems of the conventional and autonomous vehicles are disclosed (2). In the paper sensor spoofing against GPS, LiDAR (LIght Detection and Ranging), and other sensors can confuse the conventional and autonomous vehicles (2-5). However, the described security vulnerabilities problems against sensors are all solvable as long as the facts are disclosed. Recently, LTE/3G hijacking methods have been disclosed (6). They may cause more severe problems in our society from the viewpoint of security. In order to achieve safe/reliable vehicles, we must regulate the security standard for the conventional and autonomous vehicles. Remember that the conventional vehicles with CAN bus have no security embedded (2-5). If we neglect the facts and the security problems in our vehicles, we will not have the happy future.

    References:
    1. Joan Claybrook et al., “Autonomous vehicles: No driver…no regulation?” Science 06 Jul 2018: Vol. 361, Issue 6397, pp. 36-37
    2. Y. Takefuji, “Connected vehicle security vulnerabilities”, IEEE Technology and Society, pp15-18, March 2018
    3....

    Show More
    Competing Interests: None declared.