Brad Templeton Home


Robocars
Main Page


Brad Ideas
(My Blog)


Robocar Blog


My Forbes Articles


(Book from 2008)


The Case

Roadmap

Roadblocks

Car Design

Problems

Privacy

Stories

Deliverbots

Whistlecars

Future Transit

Objections

Myths

Geeks

Competition

Parking

PRT

Accidents

Notes

Teams

Government

Regulation

Motornet

Urban Planning

Development

Congestion

End of Transit

Definition

Statistics

Glossary


Sidebars: Charging

Speed Limits

NHTSA Levels

Valley of Danger

Simulator

Google Cars

Robocar Oriented Development and the New City

Robocar Oriented Development and the New City

This essay examines the potential future big changes to cities as a result of robocars. Most of these changes will be inspired by one key element of robocars: because they can drop passengers off, and then go do other work or park themselves densely in more remote lots, the need for large amounts of parking surrounding commercial buildings should diminish greatly, particularly in suburbs and non-central urban areas. If the land devoted to parking can be repurposed, what does that mean for the city?


One of the most fundamental purposes of the city is transportation. Some would even argue that other than the value of shared infrastructure, the only real purpose of the city is to reduce travel times and costs to the places you want to go -- work, friends, school and shopping -- you like it when those things are close.

Because robocars will change transportation so much they will also change cities. The car itself caused cities to change in the 20th century and drove suburbanization and other trends.

It's hard to predict exactly what the changes wrought by the robocar will be. People may tolerate longer commutes or more congestion. They may be happy with a school that's distant from their home, or having networks of friends, their own and even their kids', that are scattered around the city. They may instead crave short, predictable travel times to a wide range of destinations. They may rarely walk anymore, or they may use the robocar to take them more conveniently to interesting places to walk.

With so many changes, I have issued a call for urban geographers to begin research into the area. I have a short summary article on urban planning issues that lists many of the potential effects of robocars, in that we might understand where they will lead.

Though uncertainty is high, we can still speculate on what those consequences will be. Some of the possible consequences are viewed negatively -- increased sprawl, increased vehicle miles, enclave-style living and reduced walkability.

There is an alternate possibility for increased walkability that comes from a robocar's ability to drop you off at a destination and go park/store itself or do other work until it or another car comes to pick you up. Large sections of our cities are devoted to parking today, as well as the sides of most of our streets. We demand free parking anywhere that there is not a serious under-supply of parking, and this free parking supply subsidizes the private car.

This is particularly obvious in the suburbs. The retail streets of suburbs feature vast amounts of parking, often legally required by the building codes. As such they are not dense, and it is not expected or pleasant to walk or stroll from one strip mall to another.
Typical suburban sprawl, buildings set back from street, surrounded by parking

Shopping malls have a pleasant, if sanitized, all-weather area for strolling and shopping, but they are surrounded by vast fields of parking -- walking to or from the suburban mall is very rare. These parking lots tend to be sized for the peak demand, which is typically the holiday shopping season. Much of the time they sit at low utilization.

Denser urban retail streets feature more closely packed shops with less parking and often no free parking. They are harder to drive to (which put them at a competitive disadvantage to malls and suburban streets) but once entered, walking between shops and even strolling is the norm. They are also typically embedded within denser residential areas, and people who live in those can walk to the retail zone. (In some cases it is a fairly long walk, and as such the houses that are close to the retail block command a significant premium -- though there is a trade-off between the easy walk and the increased bustle.)

When these urban clusters are associated with a stop on a major transit line, this is often called a Transit Oriented Development (TOD) and encouraging and designing TOD has become a theme among urban planners. Even without transit, there has been a push in some lower density areas to produce denser neighbourhoods that mix retail and residential (with homes over stores as used to be traditional or condo towers) to restore the active urban environment. These tend to be upscale and curated, discouraging other elements of urban density that are blamed for the flight to the suburbs.

Many people, especially families with children, also still value the lower density of suburbs. They demand larger lots and larger houses, want space between the houses, and insist on safe, low-traffic environments for their children to play and socialize. They often don't let the kids leave their block until they are much older. This in turn has led to the spread-out parking-surrounded strip-mall style of lower density living.

Consider the concept of "Robocar Oriented Development" with many parallels to TOD. In particular, consider retail streets which have minimal parking. On these streets the shops and restaurants are more tightly packed, like a traditional urban street or outdoor version of a suburban mall. Instead of being set back far from the sidewalk, the shop entrances are near it. The street is potentially narrower, though it features space for convenient stops by robocars for pick-up and drop-off without blocking traffic.

As an alternative there may be alleys behind the shops where pick-up and drop-off take place. Indeed, there are number of smaller town urban spaces that have developed this way to avoid heavy "looking for parking" traffic on the main street. Such spaces have placed lots behind the stores or built multi-level parking structures. An example is the suburban downtown of Mountain View.

In the robocar world, many people have arrived in robotaxis. These zoom off to other fares and need minimal local parking. Some have arrived in personal robocars and want their vehicle to wait nearby for them. These vehicles could stay in parking lots a modest distance from the main street, but packed "valet" style using 40% of the land of conventional parking. If density is low, such land will be readily available. If land is expensive, parking structures may make the most sense, or vehicles may travel further to places where land is not so expensive, requiring a few minutes more advance notice when people are leaving.

It should be noted that robocar style parking allows an urban retail area to share its entire parking capacity. Today, because people won't walk a long way from their parking spot, each store has to have enough parking close to it. Malls are actually more efficient at this, sharing all their parking. Suburban streets however tend to give each retailer enough parking and don't expect a shopper to park in the lot a (long) block away.

Robocars can also encourage walking by allowing one-way trips. You might find yourself dropped off for a trip at one end of an urban street, then strolling to the other end while shopping or doing errands. At the other end, you do not need to walk back to where you were dropped off, you can walk forward and be picked up at the end of your journey. You might also have brief visits with your car at midpoints along the stroll to let it hold packages.

The ROD urban center could cater to anybody, but it will be particularly popular with people who live near it. As before, some will live close enough to easily walk to the street, but a much larger set of people could also take advantage of the street with quick on-demand robocar service. Living a mile away (which is pretty common in low-density towns) would mean a trip of just 2-3 minutes, which is shorter than the typical walking time of people living in expensive homes that realtors describe as "Steps to the retail street." If service is fast the mile-away house may be as convenient to the street as a house 1/5th a mile away. That's good for the homes and also for the retailers.

The benefits for the elderly and disabled should also be considered as they become a larger fraction of society. These folks not only can't walk very far to get to an urban street, they may not be able to stroll it easily. They will be able to reach such streets more easily, and if need be, have a vehicle take them on short hops, even just down a block.

It should also be noted that the huge reduction in needed parking that comes from Robocars and Robotaxis frees up land in all but the densest cities. Robotaxis and shared-cars don't park much at all, causing a drastic reduction in vehicles per person in the city, and private robocars could still park "valet" style in existing lots, or even double-park on the sides of streets. These old parking lots could become parkland or new development according to what is desired in the city plan.

Residential density itself may increase as some houses eschew driveways and garages. While car owners do like to protect their vehicles, and they also like the immediate access that comes with a vehicle being on-site, the huge cost of garages and driveways in terms of land is hard to ignore. In many suburban developments, houses have garages taking up a large portion of their street frontage, and the house is pushed even further back from the street. The front yard is rarely used and the main rooms of the house focus on the back yard. If this is seen as a cost it will be resisted.

Even with garages and driveways removed or reduced, people still like larger yards and space between neighbours, so suburban density probably won't drop to dense urban levels. In fact, owners may just be glad to have more land for other purposes. Tight packing in high density areas lets more people live within a short walk of the urban street -- but this becomes less valuable with quick robocar transporation to get there, and of course those not wanting such trips will want low density. If land is expensive and city plans allow it, higher-density packing of no-garage houses will probably arise.

Rural ROD

Robocars may also be a means to promote lower density lifestyles where land is cheap (or owners are very wealthy.) There are many who like a rural lifestyle where they don't see the neighbours and accept it's a trip to get anything or anywhere. The robocar makes this easier too. First it makes the long commute more tolerable -- possibly done sleeping -- and even local trips are better. Children could also be taken to school by robocar, though trips would probably have to be much shorter than adults would tolerate for commutes. Children probably would ride together in 4-6 person cars akin to a school bus, as they don't get to choose their transportation and might enjoy the social aspect, where adults are less fond of riding with strangers or going out of their way.

The Deliverbot concept also offers the ability to order anything off the web and get very quick delivery -- even perishables. This will come at a higher cost than in the city as the trip may be a few miles, but small delivery robots could be light and efficient.

Small rural towns will continue to exist. They rarely had parking problems unless they had tourists, and the parking need will only drop.

Robotaxi service will exist in rural areas but probably won't offer quick spontaneous pick-up. Robocar ownership will be more common, and the land is plentiful there to store the vehicles. Robotaxi service would still work just fine for scheduled trips, and might be quite suitable as an alternative to a second car, or cars for teens.

Will we see Robocar Oriented Development?

Some decades from now, when most cars are Robocars, it seems unlikely that new business destinations would want to devote large amounts of their land to parking, nor does it seem that cities would force them to do so.

In addition, over the first decade of robocars, existing properties will see their parking lots get lower utilization. If land grows in value, there will be pressure to develop ("infill") this lot space into additional retail space, and possibly to develop to the sidewalk if sidewalk traffic is increasing. Some parking will remain for human driven vehicles, and storage of owned robocars attending on their owners. There is an attaction in having your robocar wait at the store in which you are shopping -- you don't have to remember to summon it, just walking out of the store will be enough for it to be there for you immediately. But as more customers use robotaxis the spaces will be vacant and calling out for other use.

For retail trips, it is likely that one's mobile device will be the method of payment. That provides a nice cue to the car that the owner is about to need it without any action by the owner, and this allows cars to store themselves a minute away and still be at the door on time.

Brand new developments will also consider ROD approaches, or may be required to by city planners. In addition, TOD zones may be readily converted to mixed TOD/ROD style (though they might not have the pick-up/drop-off zones placed so conveniently. If the transit at a TOD station is faster than road travel to another location due to congestion, we may see a lot of people desiring robocar trips to the transit station, but those cars will not need to wait there as such trips tend to be longer ones.

What happens to Downtown?

Downtowns, also known as Central Business Districts (CBDs) have been the heart and core of cities. In the CBD you find the central transit hub, and it's by far the easiest place to get to on transit, sometimes the only easy place to get to. You have skyscrapers and expensive parking in structures rather than lots. Depending on the city nobody lives there and it's a ghost town of concrete canyons outside of business hours or there's lots of residential property and a lively scene. In the late 20th century, cities shifted between these modes.

Before web shopping, downtowns (or the lower-rent portions of them) also were the key place for the specialty shops that could not support themselves outside of a big city. The rich corporations like banks are there to be showcased, and while parking is expensive and traffic a nightmare, there are lots of services for employees, and if you visit other companies in the CBD as part of your business day you can often walk. Big non-office destinations are often found in the CBD, including cinemas, theatres and stadiums. Convention centers, tourist destinations and fancy hotels are often there as well. Local government is usually there.

The robocar will make us re-examine why everything that's in a CBD is there. There are pressures that push things away from CBDs -- high rents, expensive parking and heavy traffic, as well as crime and homelessnes in some cities. Other things pull you in -- the vibe, walkability, the high road capacity outside of rush hour, synergies with other businesses, cachet and easy transit access.

Robocars could remove parking issues, and mitigate traffic issues to various degrees (you could, for example, be productive while stuck in traffic.) They will downplay the value of convenient transit access, making vastly more locations just as reachable by robotaxi. "Destinations" like stadiums, convention centers and cinemas may find limited synergy with other things in the CBD and find less value there. If you're going to a movie you may just prefer a convenient cinema close to you and your friends. Eventually robocars might eliminate a lot of congestion, making the CBD more attractive.

Hotel/Convention business is harder to predict. With convenient access between airports and CBDs, airport convention centers may make more sense. Many people want to stay in the heart of a city -- tourists certainly should, since tourist destinations will likely remain in downtowns, especially in older cities. However, it's also possible that tourists might greatly prefer a $100/night room with a cheap 10 minute robocar ride to their true destination over a $300/night room in the CBD.

Already may cities, particularly newer ones, are becoming "polycentric" with many smaller business districts. By making it easier to move at will among the zones, robocars may encourage that. Polycentrism has efficiencies, since it's fairly inefficient to have people go all the way to the CBD for many services in a spread-out city.

The Deliverbots

The nature of urban retailing may be changed by Deliverbots. Delivery robots could offer the option to get almost any product (and many foods) at big-warehouse prices with sub-30 minute delivery. Local retailers will not be able to compete with this for many types of products -- those that have been falling to the big-box stores will fall further. As a result, local retailers might become mostly the type that people prefer to shop at in person. Highly specialized stores, and stores for products that you want to see and try on before buying. Restaurants which focus on ambiance and socialization can thrive, as can places which get impulse customers walking down the street.

The local retailers could also make use of the deliverbots, in this case offering 3 minute delivery instead of 30 minute delivery. But the premium can only be so high and only some products will be practical here, including cooked-to-order and fresh foods, small tools and parts and so on.

To be a fit for such a retail street, you need products and services which are not "destinations" but which people like to view as one component of an outing. So a dentist or lawyer might not locate here (except because they like it themselves) but a spa might. I expect retailers to adapt and come up with new business models we have not yet seen for retailing in the instant-delivery and easy travel world.

I also expect deliverbots to change the restaurant world. They will allow "home restaurants" where skilled chefs make food using just-in-time robot delivered ingredients for people within a few miles of the kitchen. Food might be scheduled in advance and delivered while hot. A meal might come from 4 different chefs, each making a large quantity of tonight's special -- highly efficient with no wasted chef time and no wasted food, and modest electricity for the robots.

With the at-home catered meal becoming cheap and high quality, the sit-down restaurant will probably focus on service and atmosphere. As such it will want to be on a busy street. In many parts of the world all dining in good weather is done on the sidewalk, not indoors.

To Rule the Mall

It's harder to predict the future of today's mega-mall. Robocars will make it easy to go to the mall, as they make it easy to go anywhere, and they'll drop you off right at the door or even inside the mall, though not right at the store's door as they could do in a ROD street. The mall's stores would suffer the same competition from warehouses with 30 minute robot delivery. A ROD street could be a "pop over" destination for those within 1-2 miles, while a mall might be a "real trip" with longer travel times -- but the time will be productive since you aren't driving.

Malls offer climate control which will be a huge factor in places with cold, hot or rainy climates. ROD streets could in theory start enclosing, but that makes them hard to tell from malls. Malls are curated to certain styles and have security to keep out the homeless and others they don't desire; ROD streets will be more organic and under city rules.

We will probably see innovators design entirely different retail and social spaces in the robocar world.

It's all about the Children

I believe the core driver of urbanism and suburbanism is children. When children are born, their parents shift their priorities immensely to meet the needs of the children. They perceive strong needs for space, safety and the best schools for their children. Recently, parents have become even more protective, driving their children to school even when the walk is modest; discouraging the children from going out on their own in urban spaces even though crime rates are in steady decline.

While robocars suitable for use by small children will be accepted later than robocars for teens and robocars for adults, when they do come they will alter things considerably. It could be possible to have a child attend a school many miles from home with easy travel for the child and little inconvenience for the parents, at least above a certain age. Depending on how school is paid for (and what schools you are allowed to send your children to using public funds) this can allow all sorts of new things without friction -- school "shopping" and even a movement away from the neighbourhood school.

Children might also have friends from all over -- thanks to the internet they already do -- and still be able to socialize with them, though not as spontaneously as might come from meeting them in the park or walking home from school with them. For some families, walking home from school will still be seen as positive. (Rural children have had such an environment for a long time.)

Parents like larger houses and yards for their children, and they like enclaves with low perceived crime and few lower income or homeless people. Childless adults instead tend to prefer more interesting and walkable neighbourhoods.

Robocars also make it easier to arrange babysitting, to send children to places where they can be supervised, or to share care duties among an extended family or group of trusted friends. (This assumes a car that is trusted to move a smaller child under remote video supervision in a vehicle the child can't leave except in an emergency. That may take some work to produce at a level it will be trusted.)

As people's ability to satisfy the perceived needs of their children changes, so will neighbourhoods and cities.

This remains very complex

As noted, predicting urban trends is difficult, both for technologists and urban planners. Comments are welcome in the blog thread.