The Law Will Be Fixed

September 22, 2017

In the first part of  my recent article about calls for legislation in response to the Alliston case I said the following:

“These calls, or perhaps more accurately the potential legislation to which they relate, can be placed into two categories: those which relate specifically to undesirable manners of cycling, and those which relate to all cycling.”

Sure enough, yesterday the government announced an “urgent review into cycle safety”, comprising two phases which precisely match those two categories.

Make no mistake: the effect of Phase One will not be a transformation of the level of safety on our roads, nor even a transformation of the legal system’s punitive power. A cycling homicide law would be of use roughly once a year and sentencing would not in practice be much different from what is applied by way of existing offences. The real issue is not the proposed legislation itself, flawed and largely redundant though it may be, but the wider process into which it fits.

The effect of Phase One, which is a discussion of bad cyclists, will be to frame the discussion for Phase Two, which is a discussion of all cyclists. I hardly need to explain the mechanisms of influence there, but they certainly are there—regardless of whether you believe them to be there by intent or by accident.

This article, then, is a discussion of Phase Two. And to frame this discussion, let’s talk about autonomous vehicles.

The Carrot

The market for autonomous vehicles will be big. Properly big.

A recent report published on the UK Government website predicts that in under 20 years’ time the global market will be worth £907bn. Over a trillion dollars. That’s roughly the current size of the entire economy of Australia, the 19th largest national economy in the world.


And that’s just vehicle sales. Never mind the employment costs of drivers.

Take just one example, News UK, the part of News Corp which publishes its UK newspapers such as The Times and The Sun. A back-of-envelope estimate (based on daily delivery vehicle usage and average salaries for HGV drivers and hourly rates for van drivers) suggests that it stands to save something approaching £10m per annum on distribution driver pay alone once its fleet of over 400 vehicles is autonomous.

And then there’s effects even beyond that: the drinks industry is popping corks in anticipation of sales heading north.

But back to that trillion dollar prize for the automotive industry. Before it can get its hands on the cash, it has a few problems to solve. Engineering problems, which demand costly investment and delay delivery; and legal problems, which create risk and complicate the delivery process.

How do you detect humans in the dark? How do you write an algorithm that knows how to safely pass a cyclist in an urban area thick with traffic and junctions? What are the repercussions to the business if one of your cars kills someone? The common feature of many of the most difficult questions is a human third party with no hard engineering accompanying them. With no car, they’re hard to detect, their movement is volatile, and if your car hits them things tend to go a bit sour for all parties.

But it doesn’t take engineering know-how to realise that the easiest and quickest—and hence the most economically astute—means of dealing with a problem is to eliminate it. At a pure economic level, it makes no sense to spend time and money engineering a complex and error-prone solution to detect people, when you can instead have them make themselves easy to detect, or even have them moved out of the way altogether.

The Model

From both an engineering and a legal perspective, there already exists a model which broadly represents the ideal. It’s a model which enabled autonomous vehicles to be active over half a century ago. It is a railway.

The crucial aspect of a railway from the point of view of autonomy is not that trains run on rails, even though that does eliminate some problems early technology would have struggled with. The crucial aspect is that trains can assume their path to be clear. Railways are off-limits to the public: stray onto them and the risk of being hit is yours, in both practical and legal terms.

Of course, any attempt to turn roads into pure domains of autonomous vehicles—a process for which I’ll coin the phrase “modal cleansing”—would have only limited success, but the underlying point is this:

To solve the problems of autonomous vehicles one must not only control the vehicle, one must control the system.

The Technology

When it comes to the roads, the technology which underpins control of the system—or at least the engineering aspects of that control—is called V2X communication (short for “vehicle to everything”). This comprises subcategories such as V2V, V2P and V2C (vehicle to vehicle, pedestrian and cyclist respectively). It also allows vehicles to communicate with the infrastructure itself: traffic lights, pedestrian crossings and so on. Multiple V2X technologies exist, using wifi or cellular networks to transmit data.

The key feature of V2X is the degree to which it can be replied upon, because it allows computers to communicate in natural formats. Vehicle-isolated technologies such as Lidar and cameras have to read the physical world, but network-based V2X technology allows cars to read data. Instead of spending time analysing a series of images looking for people on bicycles, with a less than perfect rate of success, a car can simply make requests for devices in its vicinity and within milliseconds understand its environment.

V2X has been in development for around twenty years, and V2P/V2C trials over cellular networks are already underway in Australia. As Telstra’s CTO puts it, actively tracking pedestrians and cyclists marks “an important step in the journey to fully autonomous vehicles”.

The auto industry has quite possibly settled on V2X as the main engineering solution because of its ability to simplify the problem domain. Lidar and cameras will supplement this, but by their nature they will be secondary technologies, designed to fill in the gaps to a reasonable (read economical) extent.

But the auto industry needs a platform for V2X adoption beyond car sales. V2V is easy to roll out, of course, but V2P and V2C aren’t. How best to achieve that?

The Prediction

If I were solely interested in maximising the economic welfare of an autonomous vehicle manufacturer, and had no interest in public liberty or the benefits of active transport, this is what I’d do.

V2X means there is a need to attach some form of connected electronic device to you. This may be a smartphone or it may be a tag, but it will communicate your position to vehicles and infrastructure around you. In the case of cyclists, a good way to ensure everyone has a tag is to ensure everyone has something within which a tag can be embedded. That means a helmet, a hi-viz vest or a registration plate on the bicycle. Of these, two have distinct advantages to the auto industry: helmet laws are known to dramatically reduce the number of people who cycle, and the reflective material on hi-viz jackets makes the wearer easier to detect by Lidar.

So, expect to see calls for a cycling test (which will almost certainly help reduce cyclist numbers further) for which you earn a licence number, combined with either mandatory helmets or mandatory hi-viz vests to which that licence number and its associated V2C tag will be attached. Anyone who knows the first thing about negotiation will see the obvious tactic of calling for both helmets and hi-viz, and  then “compromising” on just one—and I suspect it doesn’t really matter which.

What about children? Easy. Ban under 10s from the road: they’re below the age of legal responsibility and surely no sane parent would let their kids cycle in traffic anyway. This could be eased by merging existing law with Home Office guidance, and allowing them to ride on pavements.

Then there’s the issue of pedestrians. It’s politically nigh on impossible to insist on a dress code for pedestrians (in urban areas, anyway—those outside of streetlit areas might take a more pessimistic view) so different tactics are required. A V2X app on a smartphone is easy enough, but people will still try to roam freely. So a jaywalking law will cover off those tricky incidents where cars fail to detect people in unexpected places: if the worst happens, at least the vehicle maker won’t be liable. Cross at the V2X-enabled crossing, or on your head be it.

So here’s my prediction—not a paranoid reading of tea-leaves but a rational conclusion based purely on economics and engineering, and to hell with morality or the greater public good: Age-restricted cycle testing and licensing, one piece of mandatory cycling wear, and the jaywalking law. You’ll see calls for much more (negotiations have to start somewhere, so it’s best to start with more than you want to settle on) but I suspect these are the core goals.

None of these things are truly about safety or responsibility—the rational arguments about where road danger actually comes from and how regulation hasn’t suppressed it will be actively drowned out—they are about control.

Personally, while I appreciate the engineering benefits of V2X, I loathe the thought of being forced to dress like a construction site worker or to take unreasonably convoluted routes to walk across roads, and I resent the looming step-change in the domination of motor vehicles at the expense of human-friendly places. If that’s an opinion you share then you’d better stand up right now, because this battle was planned—and probably largely fought—long ago.

The question which would have already faced the auto industry is not so much one of what to achieve, but of how to achieve it.

But that’s already been answered, too.

The History

Back in the early 20th century streets were for people to roam freely, but the transformation from the social view of the car as “Satan’s killing machine” to where we are today was wholly manufactured in the 1920s by motoring organisations and car manufacturers—a process famously documented in Peter Norton’s book “Fighting Traffic” and, as a result, widely reported by many outlets including the Washington Post.

Even though techniques of influence in the 1920s were not subtle by today’s standards, the lobby knew where best to channel its efforts: schools, governments, the police and the media all played their part in the motor lobby’s complete overhaul of public attitudes, all complicit in laundering the pro-car, anti-pedestrian propaganda.

From Citylab“AAA and other auto clubs turned first to the younger generation, financing safety education programs in the public schools that were designed to teach children that streets are for cars, not for kids. They funded safety patrols that taught kids they had to stop for traffic, not the other way around.” It’s still going on under your nose. The cars-not-kids view is so entrenched that government messaging is fully in tune with the 1920s motor lobby.

I’ll leave you to dig out more information around this—there’s plenty of it—but this paragraph from Vox’s article about Norton’s book should give you pause for thought:

“Before formal traffic laws were put in place, judges typically ruled that in any collision, [the driver of] the larger vehicle—that is, the car—was to blame. In most pedestrian deaths, drivers were charged with manslaughter regardless of the circumstances of the accident.”

Back in the 1920s, manslaughter laws sustained the prevailing view: that people should be able to roam their urban environment freely, and that those in charge of vehicles should take responsibility for them. The introduction of traffic laws served only to upend this world view; they were not created to protect those outside of cars, but were part and parcel of the handing over of streets to the motor car, banishing those without cars to the sidewalks and crosswalks by way of jaywalking laws.

Yet now we bear witness to a campaign demanding more traffic law in response to the death of someone stepping into traffic away from a crossing.

History tells us that is, for want of a less ironic phrase, a step in the wrong direction. Its inevitable gift to the motor lobby is illustrated by another collision involving a pedestrian and a cyclist. On this occasion it was the cyclist, Ben Pedley, who lost his life. His brother says he “has launched a campaign to have the law changed”.

So, from two almost identical events, we have two perfectly opposite campaigns to change the law: one to crack down on cyclists, one to crack down on pedestrians. Suddenly it seems as if, unlike a century ago, the auto industry can almost sit back and do nothing: the groups it wants out of the way are so busy kicking each other out of the road that they can’t see the headlights bearing down on them.

The public anger at these rare incidents, the desperate grabbing for legislation to somehow make everything safe, will continue while five people a day silently die in motor vehicle collisions, and the motor industry will ride this wave of public sentiment all the way to the statute books.

Have you noticed that the government has dragged its heels over a promised review of road law for the past few years, yet after a few weeks of persistent media anti-cycling messaging that review has become “urgent”? I know, I know—best not think about it.

Phase Two will soon be upon us, and Phase One will obliterate both the lessons of the 1920s and the wider context of the overwhelming cause of not only death and injury on the roads, but of human-unfriendly urban environments: so effective is motor traffic at ruining people’s lives that it has been used explicitly to that end.

The Wager

Do you want to bet that the proven strategy of the 1920s, the sort of campaign that seismically shifts public attitudes for the next century, isn’t being employed again?

Because, before you do, remember that you’re betting against a trillion dollars.

You’re up against big money here. Good luck with that.


Premier Partners

Categorised as:

Bez Opinion