Planning for Autonomous Vehicles

I’ve been thinking a lot about how autonomous vehicles (AV’s) will change how we plan our roads lately. In the back of my mind, there is always the broad thought that AV’s will change the world for the better, but will they? I’m starting to have my doubts…

Defining AV’s

The Society of Automotive Engineers (SAE) provides six levels to classify autonomous vehicles:

  • Level 0: Automated system issues warnings but has no vehicle control.
  • Level 1 (”hands on”): Driver and automated system shares control over the vehicle. An example would be Adaptive Cruise Control (ACC) where the driver controls steering and the automated system controls speed. Using Parking Assistance, steering is automated while speed is manual. The driver must be ready to retake full control at any time. Lane Keeping Assistance (LKA) Type II is a further example of level 1 self driving.
  • Level 2 (”hands off”): The automated system takes full control of the vehicle (accelerating, braking, and steering). The driver must monitor the driving and be prepared to immediately intervene at any time if the automated system fails to respond properly. The shorthand ”hands off” is not meant to be taken literally. In fact, contact between hand and wheel is often mandatory during SAE 2 driving, to confirm that the driver is ready to intervene.
  • Level 3 (”eyes off”): The driver can safely turn their attention away from the driving tasks, e.g. the driver can text or watch a movie. The vehicle will handle situations that call for an immediate response, like emergency braking. The driver must still be prepared to intervene within some limited time, specified by the manufacturer, when called upon by the vehicle to do so.
  • Level 4 (”mind off”): As level 3, but no driver attention is ever required for safety, i.e. the driver may safely go to sleep or leave the driver’s seat. Self driving is supported only in limited areas (geofenced) or under special circumstances, like traffic jams. Outside of these areas or circumstances, the vehicle must be able to safely abort the trip, i.e. park the car, if the driver does not retake control.
  • Level 5 (”wheel optional”): No human intervention is required. An example would be a robot taxi.

Level 3 cars are beginning to be presented by manufacturers now (Mid 2017) for sale next year. At the rate technology advances, I’m going to predict Level 4 could be available in the next 5 to 10 years, and Level 5 in the next 10 to 20 years, well within current planning horizons.


I keep coming back to the fundamental principle that people have to get to where they have to get to. Maybe this is looking at it too simplistically. We all have to live somewhere and we all have to work somewhere just like we do now. Whether the population as a whole does that by driving themselves like they do now, or they summon a self-driving car to do the driving, thats still the same volume of traffic on the roadway. Unless we all start working from home, trips should remain similar or vary in line with population growth.

Impact: Car traffic unchanged

AV’s could create more traffic, with the requirement to actually drive the vehicle removed, some that currently bike or bus may also choose to summon a car to their door as its more convenient. Those currently too young to drive or too old to safely control a vehicle, or those that never learned to drive will also now have a car based option. This is great for mobility, less so for congestion.

Impact: car traffic increases

​While its often said that the best transportation plan is a good land use plan, while house prices or desire for a single family home push people out to the suburbs, the long commute can also keep people closer to the city. Removing the need to do the driving, removing the stress that causes, allowing you to sleep or read the news or check emails while being driven, makes a long commute much less of a deal breaker. With more people choosing to live where conventional transit doesn’t work and choosing to be driven to work, we may have a far greater problem on our hands than we think.

Impact: car traffic increases

Whether privately owned and shared amongst the family, or shared amongst the wider population, AV’s will result in zero occupancy car trips, trips with no purpose other than to pick someone up, in a privately owned scenario these trips could be significant, imagine a scenario, where the car drivers the husband to work at 7am, drives home alone, picks up the wife and drivers her to work, drives home alone, picks up the teenage kids and takes them to school, drives home alone, we’ve saved on the number of vehicles but dramatically increased the number of vehicle kilometres travelled. Imagine where half the vehicles on the road are simply going to pick someone up. Where the cars are shared amongst the population, the number of zero occupancy trips should be less and of lower distance, as they can be optimized to make their next pick-up close to the last drop-off.

Impact: car traffic increases

There is potential from some slight increases in efficiency, currently, with a human driver, it is recommended that you leave about a 2 second gap from the car in front, this allows space to react to conditions ahead. With a 2 second headway, one lane has a capacity of 1,800 vehicles per hour. With connected vehicles if that safe headway can be reduced to 1 or 0.5 seconds, capacity could increase dramatically. A 0.5 second headway would give a single lane capacity of 7,200 vehicle per lane.

Impact: road network can accommodate much more traffic.


There is a huge potential for AV’s to reduce current parking requirements. If we continue to own our AV’s similar to today, we can likely reduce the number of cars per household. Where a family owns two cars, drive to work separately, while there might still be some juggling of start/end times, its far easier to make all your trips with one vehicle, when it can be summoned with a moments notice. Clearly under this ownership model the car still has to park somewhere when not in use, but we’ve potentially reduce parking demand by 50% if every two car family reduced to one.

Impact: for multi-car families, parking requires could reduce from 2, 3 or 4 spaces to 1.

Ideally, personal ownership would be greatly reduced and most would opt for a shared vehicle model. This would remove parking requirements greatly, cars would have much less down time, and the time to summon a vehicle would be much shorter as you would select the nearest available. Depending upon the efficiency of the system and the supply/demand balance, there would still have to be some parking provision for vehicles to wait, recharge, or come in for service, but as a shared use vehicle would be far more productive with much less idle time, one vehicle would likely do the work of 10 or 20 regular vehicles throughout the day, thus parking could likely be reduced to say 10% of the current provision.

Impact: most residential, commercial and workplace parking could be removed, some parking would be required to allow vehicles to return for charging or servicing, or during quiet times, this should be distributed evenly across the city.

Road Design

In a world where we have nothing but AV’s there are theoretically significant changes we can make to our current road designs, to save cost, save space, or make space available for others. In theory with connected vehicles and high levels of sensory accuracy, cars could travel much closer together, imagine if all those 3.5m lanes can suddenly be just 2m or maybe 2.6 for larger vehicles.

Impact: More car lanes in the same space, or same number of car lanes in a smaller space with remaining space allocated to walking, cycling or transit.

Car Design

Perhaps car design can also offer space saving benefits. Today we buy a car that works for us in most situations, so something that can carry our family or friends, manage a trip to the hardware store, etc, yet most travel to work on our own, not utilizing that space. If AV’s evolve to a state where we just subscribe to a shared use fleet, and assuming we still don’t want to sit beside a complete stranger, perhaps there can be a large fleet of single occupancy AV’s, and then a smaller fleet with higher people capacity or load carrying capability.

​Impact: People driving more appropriate vehicles for their trip purpose, lower energy requirements, less road space required.


Safety is one of the most obvious benefits, assuming they figure out the software and sensory capabilities of the cars. There is of course dangers of software crashes and hacking, but hopefully the benefits would far outweigh such potential incidents.There is the other issue of allowing algorithms to decide who lives and who dies should the car get into a situation where it has to make a decision. Say a pedestrian walks out and the only way to avoid them is to crash into a wall, what does the AV do? There’s likely a million potential scenarios that could play out. Would pedestrians jay walk more knowing the car is programmed not to hit them?

Impact: reduction in collisions will prevent deaths and reduce disruption to traffic flow.

Back to Planning

I once thought it was dangerous to build capacity because it won’t be needed. That AV’s will save the day and reduce our infrastructure needs, but now i’m not so sure thats how this will all play out, as there are many reasons traffic volumes will increase. Smaller vehicles with smaller gaps have the potential to claw back capacity to accommodate that increase in volumes.One thing that is certain is that parking requirements will almost certainly reduce. In terms of surface parking lots, freeing up land for residential, commercial or employment uses in our urban centres is definitely a good thing. Some on-street parking will still be required, but for drop-off and pick-up purposes, but perhaps in limited locations freeing up space for pedestrian, bike or transit facilities.Regardless of the above, we must still create streets we want to be in, places for people rather than for cars, providing car capacity above providing a place for people is rarely a good decision.


There are many different ideas on timelines for these changes to take effect. Cars are only now being offered with Level 3 Autonomy. Level 4 and 5 are likely 5 to 20 years off, but certainly are coming within current planning horizons. Level 5 is where we will begin to see significant disruption to existing travel patterns.For much longer than that, we will see a mix of human controlled vehicles and varying levels of autonomous vehicles. This transition with multiple vehicle types will prevent the significant efficiency benefits from being realized any time soon, unless we have AV specific lanes or streets.

So within current planning horizons, we can expect slight increases in traffic towards the end due to trips by current non-drivers and the beginning of zero occupant trips and only once we can have zero occupant trips, will we reduce parking requirements. This is demonstrated in the chart below.

Funny thing, autonomous vehicles are essentially taxi’s with better driving, collision avoidance and less fear for your personal safety. We could be reaping the reduced car ownership and parking benefits right now if the taxi industry could offer competitive costs, reliable pick-up, and personal safety. Uber and lyft somewhat took up that mantle with app based requests and payment, the uncertainty over the driver is the only challenge…

Originally published at

Thoughts on how we move around, whether by walking or cycling, transit or automobile, and how urban design influences that.