The Columbia River Crossing bridge and freeway widening project (R.I.P. 2005 to 2013) was estimated to cost $2.8 billion, down from earlier estimates of $3.5 billion. Lack of consensus, money and vision killed the old interstate bridge plan.
Here’s my testimony for the new bridge committee, promoting the concept of EVs and tunnels in October, 2019 and a newer concept paper in October, 2020 (pdf). Tunnels may off-load I-5 traffic more cost/effectively than Bus Rapid Transit or Light Rail, the preferred alternatives.
Here a brief description of “The LOOP” alternative (pdf), for the Jan 20, 2021 I-5 Bridge meeting. Musk’s electric shuttles, running autonomously in an underground tunnel instead of Bus Rapid Transit on the I-5 Bridge would be cheaper, faster, more convenient, and less polluting. Save time and money.
The Loop is FREE in Las Vegas. Why can’t we do it here? Why subsidize TriMet or C-Tran using hundreds of millions in business taxes if there’s a better way?
The planned bridge and its 22-lane expressway would have done little to solve the main problem; too many cars. It’s axiomatic: more lanes = more traffic. The solution is to make mass transit more convenient, faster and cheaper than cars.
It’s not rocket science. It’s a problem that can be solved.
Wider highways just get filled up. More lanes = more traffic. Reducing congestion is the goal. The main problem is vehicles carry only an average of 1.2 passengers.
ChargePoint expects it will take about 20 years to replace all 250 million cars and light trucks in the United States with electric vehicles. So revenue from the Gas Tax may disappear only 10 years after the replacement I-5 bridge is finished.
A tunnel for electric vehicles may be inevitable. No pollution. Cheap. Drivers optional.
An I-5 tunnel in Seattle is proposed. It might be like their Link Light Rail tunnels, the University Link tunnel, Beacon Hill Tunnel, the Downtown Transit Tunnels, the Mount Baker tunnels, the SR-99 tunnel…the list goes on.
The tunnel option was explored in the early days of the ill-fated Columbia River Crossing project, reports The Columbian.
That proposal basically put the entire freeway underground. But tunnels are generally point to point. You don’t need to be a highway engineer to understand the problem of (somehow) connecting to State Route 14, downtown Vancouver, Hayden Island and Marine Drive. Underground cloverleafs are untenable.
Linking an 8 lane freeway tunnel under the Columbia to SR 14 and to Marine Drive, already at capacity, would be (1) an engineering nightmare and (2) do nothing to alleviate congestion on I-5.
Here’s the Tunnel Study for Max Light Rail under the Willamette River. Projected cost for the proposed 3.5 mile tunnel from Lloyd Center to Goose Hollow, with three underground stops, is $3 – $4.5 billion. It uses the $1B per mile cost modeling, similar to Seattle’s Alaska Way tunnel.
TriMet’s rationalization: the Steel Bridge simply cannot accommodate more train traffic. Plus trains can have more cars than the two, limited by 200 ft long blocks in downtown Portland. Either build another bridge or build a tunnel. A tunnel has more earthquake resiliency, says TriMet, and could accommodate longer cars than a surface train.
Competitors, such as The Boring Company, promise much cheaper tunneling. Primarily that’s because they use smaller (12 ft vs 24 ft) tunnels displacing 1/4 the volume. Boring Company tunnels also don’t need overhead power or tracks – they’re for electric cars and shuttles. Boring Company estimates their cost at $50 million/mile vs $1 Billion/mile for light rail. The same 3.5 mile tunnel, therefore, might cost less than $200 million using Hyperloop. Their one mile Las Vegas tunnel is expected to cost $48 million.
A Hyperloop tunnel, by contrast, would accommodate electric shuttles and personal electric cars — which can then use ordinary roads. Tri-Met operates mostly using taxes and bonds and is a closed system. Only TriMet can use it.
I believe The Boring Company provides a viable alternative to TriMet’s vision. But it would require grassroots support from a coalition of private enterprise, fiscal conservatives, riders and the pubic.
We offer an alternative vision, using Hyperloop instead of a new Columbia River bridge. But THIS is the slower, cheaper version of Elon Musk’s “hyperloop”. No maglev or 700 mph speeds — just a paved underground tunnel for electric vehicles. More lanes. Lower cost. No eminent domain of land for freeway construction.
Virgin Hyperloop One, an inter-city competitor to The Boring Company. It also hopes to whiz people from, for example, Dallas to Houston in 30 minutes. Proponents say the capital and operational costs of a hyperloop system could be two-thirds that of high-speed rail – and much faster. The local subway system that uses electric cars that The Boring Company has recently proposed would be much cheaper.
U.S. Transportation Secretary Elaine Chao has created a transportation technology council aimed at eliminating regulatory and legal roadblocks for Hyperloop. The US DOT aims to help hyperloop and self-driving car projects.
Many will dismiss this notion as non-sense. Our intent is to spark conversation and addresses real issues that the original tunnel proposal and the original I-5 bridge proposal did not.
➤ This tunnel concept is based on much smaller and cheaper “hyperloops”. It’s dedicated to electric shuttles. No trucks. No trains.
➤ It relieves congestion on the ENTIRE I-5 corridor to Portland, not just the bridge.
➤ It consists of two small tunnels, on the scale of 12 ft Hyperloops as planned in Las Vegas and Chicago.
➤ The first tunnel – or tube over the river – takes commuters to Expo Max Line.
➤ The second tunnel takes commuters from Expo Max to downtown.
➤ Max speed about 100 mph. It’s essentially a simple paved tunnel for electric vehicles
Hyperloop is a generic term. It’s main competitors are Virgin’s Hyperloop One and Musk’s Boring Company. Originally, the concept was for inter-city travel of several hundred miles, traveling some 600 mph in vacuum tubes. More recently, simpler, cheaper paved tunnels for electric vehicles have been proposed for local commutes.
➤ Off-load traffic and eliminate the need for a wider I-5. A tunnel between Vancouver and the Expo Center and between Expo and the Fremont Bridge could bypass I-5 congestion.
➤ Congestion is a real problem. Congestion costs Portland commuters $1.76 billion in time and fuel in 2014, according to a 2015 analysis by the Transportation Institute. Vehicle use is increasing.
➤ Make mass transit faster, cheaper and more convenient than driving. Electric transit and tunneling are enabling new approaches; reducing congestion and cost while enhancing convenience and speed.
The Regional Transportation Summit met in August 2018, to discuss a new I-5 bridge. Oregon House Speaker Tina Kotek wants to sit down with Washington lawmakers before the end of 2018 to discuss replacing the Interstate Bridge. But there’s no consensus on what’s needed or how to pay for it.
Above is a powerpoint slide-show (and pdf) that illustrates the highlights of this page. Here’s a white paper describing small, cheap bridges dedicated to autonomous vehicles, bikes and pedestrians. The intent is to siphon 10-20% of the traffic from I-5 by making mass transit faster, cheaper and more convenient.
We already have an operational North Portland tunnel. The North Portland Peninsula Railroad Tunnel has been operating daily for nearly 100 years. It’s practically under our nose. It terminates near the Fremont Bridge.
The train tunnel goes from Kenton (just off Columbia Blvd) to Swan Island, about 1 mile long and 60 feet underground. Amtrak doesn’t use it. It’s only for freight trains.
I-5 Tunnel Vision:
➤ One tunnel under the Columbia would connect Vancouver to the Expo Center in North Portland.
➤ A second tunnel, near the Expo Center, carries vehicular traffic to the current 405 bridge.
➤ Congestion caused by Marine Drive, Columbia Blvd, SR-14 and I-5 could be relieved, while transportation innovations like electric shuttles could be phased in.
The North Portland Peninsula Railroad Tunnel connects North Portland to Mock’s Bottom (below Willamette Blvd and near Swan Island).
The tunnel is owned and maintained by the Union Pacific Railroad. The concrete-lined tunnel, about 60 feet below the north Portland highlands, shortens freight movement to Washington State. A similar North Portland Tunnel for vehicles would run roughly parallel to the current RR tunnel.
The Peninsular Tunnel is used constantly, EVERY DAY. The North entrance is just south of Columbia Blvd, about 1 mile west of Kenton. The South entrance is on the lower part of the cliff overlooking Swan Island.
➤ The North Expo Tunnel offloads Vancouver traffic from SR-14 and I-5. The Vancouver terminus, near Pearson Air Park, goes under the Columbia and surfaces at Expo Center.
➤ The South Expo Tunnel offloads Oregon traffic from Marine Drive and Columbia Blvd. It connects to the Fremont (405) bridge.
North Portland Portal
The current North Portland Tunnel is about a mile long. The terminus is currently south of Columbia Blvd and east of the Sewage Treatment Plant.
South Portal to Albina Yards
The current RR tunnel comes out near Mocks Crest, in Swan Island. A parallel vehicular tunnel might be located 1/2 mile to the east and come out near the Albina RR yard.
The South tunnel might be about 2 miles long, with one terminus behind the Expo Center and the second terminus near the Albina Yards and Greeley Avenue. The Portland Albina Rail Yard Relocation Project (below) will also require more traffic capacity.
Tunnel traffic moves adjacent to Greeley Ave until it feeds into the Fremont Bridge and I-405. Terminations near Union Station or the Coliseum may be viable options.
Tunnel traffic could merge with ramps to enter and exit I-405 on the Fremont Bridge.
Rubber tired autonomous shuttles will arrive far ahead of ANY new bridge. Plan on it.
Electric shuttles can carry people to their destination more conveniently and without the expense of parking. First and last mile.
Tunnels are not new.
➤ The 2.9 mile Light Rail tunnel, west of Portland and underneath The Zoo, cost $184 million.
➤ Seattle’s 2 mile Alaskan Way Viaduct bypass opens in October, 2018. Seattle’s SR 99 Tunnel cost between $1.5-$2 billion.
➤ The 3.6 mile Transbay Tube carries Bay Area Rapid Transit (BART) trains. The tube was constructed on land, transported to the site, then submerged and lowered to the bottom.
➤ The Portland-Milwaukie Light Rail cost $1.5 billion for 7.3 miles, or $205 million per mile.
➤ Chicago selected Elon Musk’s Boring Company to design and build a high-speed underground transit system linking O’Hare to downtown. The 8 mile line would cost $1 billion or $55.5 million per mile – a far lower cost projection than most recent tunneling projects.
The original Hyperloop vision basically involves using passive magnetic levitation and propulsion to move a capsule — potentially carrying about 38 passengers — through a large windowless tube. The much shorter “loop” is a different design for transportation inside cities; basically a subway for regular electric vehicles.
Hyperloop is an open-source transportation system that uses a sealed tube with low air resistance. Musk launched The Boring Company back in 2016 to dig Hyperloop tunnels under Los Angeles and elsewhere. The Boring Company hosts its tunnel opening along with a series of product launches on December 18, 2018.
Main competitors includes Richard Branson’s Virgin Hyperloop One and Hyperloop Transportation Technologies. California’s Hyperloop Transportation Technologies has Abu Dhabi and Hamburg ports in its sights while Virgin Hyperloop One is the system favoured in the United Arab Emirates. Hyperloop One is studying 10 routes that have potential around the world. Virgin Hyperloop’s CEO Jay Walder ran subways around the world and Motivate, the bike-share company that was bought by Lyft.
High Speed Rail
Amtrak’s WA derailment in 2017 set back regional train service expansion by years. The Pacific Northwest Rail Corridor is one of eleven federally designated higher-speed rail corridors in the United States.
A Portland and Seattle trip would take 2 hours and 30 minutes at 110 mph. But the $24-$42 billion HSR system may require operating subsidies for the first 20 years or so, according to a WS-DOT report.
The California High Speed Rail is a traditional high speed rail approach, with construction plans currently on pause for an LA to San Francisco route. Too much money.
The 2018 draft Business Plan for California HSR calls for opening the initial operating segment between San Jose and Bakersfield in 2027, with the $64B first phase completed between San Francisco and Anaheim in 2033.
Hyperloop advocates argue their approach is MUCH faster for longer inter-city routes…and cheaper as well. Hyperloop would use smaller tubes and tunnels, mostly within a vacuum tube which would allow them to operate up to 700 mph. Hyperloop claims to save on land costs, sometimes utilizing the median strip on freeways for their elevated tube.
The Thing About Trains
Trains are big, expensive, slow and noisy. They require rails, overhead power lines and huge investments. They’re inflexible. Require drivers. Can’t run on regular streets.
The main problem with inter-city High Speed Rail is the $4 cost per mile per person.
Taxpayers may have to pay for 70% of the cost of HSR.
Railroad Map of Region
Autonomous Vehicle fleet operations are estimated to cost 35 cents per mile compared to the $3.50 per mile cost of a taxi—a 90 percent price reduction.
Short-hop electric Sky Taxis are anticipated to cost less than buses, rail or ferries in 10-15 years. Twice the speed, half the cost.
Electric powered E-VTOLs made by Uber, Airbus and Boeing promise to be fast, quiet, and cheap. No roadbed. No gas. No pilot. When? By 2035. Electric VTOL aircraft will likely kill HSR before it gets built. There’s virtually no infrastructure cost for EVTOLs.
Who’s going to pay for a $30 Billion HSR white elephant? You?
The Chicago (local) Loop
The 8 mile Chicago Hyperloop approach is different from the inter-city “Hyperloop”. It doesn’t operate in a vacuum. Doesn’t go 700 mph. Doesn’t use superconducting magnets. It’s basically a small subway for electric cars and shuttles.
The Chicago “Loop” would transport passengers in automated electric cars carrying 16 passengers (and their luggage) through two parallel tunnels, at speeds reaching 150 miles per hour, with pods departing as often as every 30 seconds.
The “loop” uses ordinary, rubber-tired, electric-powered cars or driverless “pods” that hold 10-16 people. Tesla has abandoned their “electric skate” plan and replaced it with autonomous electric vehicles, starting with Tesla vehicles, using ‘tracking wheels’.
The Las Vegas Loop
The Las Vegas Hyperloop would eventually cover about 18 miles and cost around $500 million. The system would likely be linked to a smartphone app, where a user could choose their desired station and a vehicle with that stop advertised on its digital signage would be dispatched to provide an express ride to the destination. Their system would feature two tunnels, running side-by-side about 30 feet below street level, each about 12 feet in diameter. It’s self-funded by Musk. No taxpayer money.
Can Musk’s Boring Company reduce the cost of tunnel boring so dramatically? Many are skeptical. Still, smaller tunnels will lower costs and investors in the Chicago/O’Hare tunnel are confident enough to put up 100% of the financing- without ANY government funding.
Proposed routes include a District of Columbia to New York run, an express service from O’Hare International Airport to downtown Chicago and a Loop in Los Angeles to transport baseball fans directly to Dodger Stadium.
Chicago’s contract with Musk for their 8 mile tunnel to O’Hare airport, 100% funded by private enterprise, would cost about $1 billion. Instead of a high-speed vacuum tube, The Boring Company proposes a low speed (150 mph) subway. But will homes and businesses have to make way for a hyperloop bypass? The validity of the Boring Company’s tunneling claims should be clear…one way or the other…in a year or two.
The original, city-to-city “Hyper”loop is tough to take seriously. Maintaining and securing a 100 mile vacuum tube with less pressure than the air at 30,000 feet would be a neat trick. How do you prevent everyone in the tube from being killed if the tube is damaged. Superconducting magnets are expensive and they require a lot of electricity. It seems like wishful thinking.
Still, a team from NASA examined the feasibility of the hyperloop, purely from a technological standpoint, and found it doable. A feasibility study by Black & Veatch concluded that Hyperloop would cost less per-ride than the cost of gas to drive the same distance.
Can Musk drill tunnels far cheaper than anyone else? How? Where’s their breakthough in drilling tunnels? They don’t have one. The Boring Company appears to rely on smaller tunnel diameters to lower cost. Half the height (12ft vs 24ft) means 1/4 the volume and cost.
Elon Musk won the Chicago Loop contract since it didn’t cost the city one dime. Private investors fronted the money at no risk to the city. Of course Chicago backed that proposal. Musk’s Chicago tunnel system claims it will be able to handle nearly 4,000 passengers an hour, about the capacity of the Max Red Line.
But BART moves 27,000 people an hour during rush hour and the interstate bridge carrys about 140,000 vehicles a day.
Building the BART line
The 3.6 mile Transbay Tube carries Bay Area Rapid Transit (BART) trains. The tube was constructed on land, transported to the site, then submerged and lowered to the bottom.
The underwater BART tube is made of 57 individual sections that were built on land at the Bethlehem Steel shipyard and towed out into the bay by a large catamaran barge. After the steel shell was completed, water-tight bulkheads were fitted and concrete was poured to form the 2.3-foot (0.70 m)-thick interior walls and track bed.
They were then floated into place (positioned above where they were to sit), the section was ballasted with 500 short tons (450 t) of gravel before being lowered into a trench packed with soft soil, mud, and gravel for leveling along the Bay’s bottom.
The Bay Area’s 2040 transportation and land use plan presumes that more people will live or work near regional transit.
For example, Plan Bay Area anticipates 250,000 new jobs, or a 40 percent increase, in areas adjacent to BART stations.
The world’s first subway opened in London in 1863 – using coal-powered steam engines. But Sprague’s electric traction motor started a revolution, and helped create the Chicago El train (1895), as well as the Boston (1897) and New York City (1904) subway systems.
A century ago, the Willamette Shore rail corridor was equipped with overhead electrical wires that supplied power to the fleet of Red Electric commuter trains that ran between Portland and Lake Oswego. Service was discontinued at the end of the 1920s, and the overhead cables were subsequently removed.
In 2014, TriMet discontinued its vintage trolley service in downtown Portland and leased two of its historic trolley cars to the Oregon Electric Railway Society to run on the Willamette Shore Line. Now that trolley may be equipped with a battery pack mounted under the floor, which could be recharged inside the trolley barn in Lake Oswego.
Cheaper than Bus or Rail?
The tunnels might be built like the Chicago Loop – with NO public funding. The 8 mile Chicago Loop plans to charge $1 per head but may cost $1 billion. That may be wildly optimistic for Portland, since the entire Max system doesn’t get 40 million riders per year.
At $1 a head over 10 years, that might total $400-$500 million – for the entire Max system. TriMet’s system is largely funded by employee taxes, of course.
The farebox covers only a small part of TriMet’s actual cost. All the rest is taxes. Buses and Max cost billions to operate.
Expo Tunnel Hub
A tunnel hub might be near The Expo Center. One tunnel goes North, under the Columbia, to a terminus near Pearson Air Park, at the junction of SR-14 & I-5.
The 2nd tunnel goes South from the Expo, merges up with the Vancouver tunnel and surfaces near the Albina train yard and an on-ramp to the Fremont Bridge. The one-way tunnels are paired. They’re only 12 ft in diameter, lowering construction costs dramatically. No trucks or trains. It won’t SOLVE congestion but it MAY be a cost/effective solution to REDUCE congestion.
Planners (and others) should study the impact of intercity Hyperloops. Professional planners have to acknowledge that intercity Hyperloop is here. It’s in the public consciousness. Portland’s transportation planners are using Sidewalk Lab’s Replica software which is powered by anonomized location data collected by millions of cellphone users.
➤ The tunnel off-loads traffic from Marine Drive, SR-14 and I-5.
➤ Electric cars and shuttles can also use today’s roads.
➤ No pollution, no gas, no trucks.
➤ No Billion dollar Interstate bridge required.
➤ Faster and Cheaper.
I’m no fan of rail. Too expensive. Too heavy. Instead, rubber tired, battery operated autonomous Pods are light and small. More flexible. Cheaper. No tracks, no overhead power, no new infrastructure, no driver. Transportation for the 21st century.
➤ Rubber-tired electric shuttles don’t need new infrastructure.
➤ They can pick you up.
➤ Vancouver to Portland in 5-10 minutes.
Luxembourg, with a population of about 560,000, has become the First Country to Offer Free Mass Transit for All. Luxembourg budgeted nearly €900 million for mass transit, but recovered only €30 million in ticket sales. Saving that overhead helped finance some of the cost of free travel. Gas for a car costs about $.12 per mile while electricity costs about $.04 per mile.
Public transit may cost an average of $7.83 per passenger who will give back only $1.26 in fares. Everything else is subsidy. A “farebox recovery ratio” can average about 15%, while the cost of collecting and managing those revenues costs nearly as much. Electric, autonomous transit may be moving towards free. Cheaper fuel, lower maintenance, no drivers.
People are paying about $2.50 a mile when they use ride-sharing companies like Uber and Lyft. If they buy a car, they are essentially paying 70 cents a mile to go the same distance. Autonomous taxis could lower that cost to 26 cents a mile, because there’s no driver to pay. That’s $.26 a mile. Essentially, free.
Advertising could cover most of the cost at $.26 a mile. Taxes and subsidizes for mass transit? Furgetaboutit!
The Last Mile
The last mile will be served by ebikes (first 15 minutes free), scooters, shuttles, or ride sharing services like Uber, Lyft and Car2Go. Scheduled or on-demand. Neighborhood electric vehicles can be summoned with an app. Convenient. Fast. Cheap.
Think it won’t happen in 5-10 years? That’s a bad bet.
Can autonomous cars provide COVID-Proof Transit? Time described COVID-19’s impact on public transit as apocalyptic. “Buses that once carried anywhere from about 50 to 100 passengers have been limited to between 12 and 18 to prevent overcrowding in response to coronavirus.”
Right-sized Columbia Bridge
With a new tunnel off-loading current I-5 traffic, a much smaller, pedestrian/transit bridge could move Vancouver commuters to the Expo Max train.
Instead of a tunnel under the river, a tube OVER the river might be an alternative choice, with a stop at the Expo Center (for connections to the Yellow Max line). After the Expo stop, it would continue to downtown Portland.
In addition to the I-5 bridge and The Tube, a pedestrian bridge could carry bikes and electric autonomous shuttles. Shop at the Jantzen Beach Mall or Delta Park. Catch the Max at Expo. The Matagarup Bridge, in Perth, is one example of a combined pedestrian/bike bridge.
The 400 ft long bridge contains an S-curved deck that leaves plenty of room for barges to pass underneath. The S curves of the bridge deck is counterbalanced by outward-leaning arches.
A much lighter pedestrian bridge, with a roll-up deck, could then be possible over the Columbia. We’d upgrade the current Interstate bridge for current car and truck traffic. Perhaps 10% of the bridge traffic (20,000 riders/day), however, may likely choose to take the hyperloop direct to the Expo Center or downtown Portland. It would be faster, cheaper, and more convenient option for a critical percentage of people.
The overall economic impact of congestion in the U.S is estimated at $2.8 trillion by 2030 – the same amount Americans collectively paid in U.S. taxes last year, according in Inrix (pdf).
Faster, Cheaper, More Convenient
More car lanes = more traffic. The solution is making mass transit, faster, cheaper and more convenient. Autonomous shuttles and electric bikes/scooters can (and will) connect the last mile. Autonomous transportation is estimated to cost 35 cents per mile per passenger vs $1 per passenger mile for less convenient buses.
Neighborhood autonomous shuttles can connect residents to mass transit. The free “Loop” (paid for with advertising), passes nearby every 30 minutes. Faster, cheaper, more convenient.
Everyone knows zero emission and autonomous vehicles are the future. There’s no question about it.
If a “Loop” and tunnels can save billions and deliver real benefits, then ODOT and PDOT officials should study it. Other cities are BUILDING hyperloops. If there’s “hype” in Hyperloop, let’s hear about it.
We need to know. We’re paying for it.
Here’s my testimony to the Joint I-5 Bridge Commission: (http://www.hayden-island.com/wp-content/uploads/2015/10/I-5-Bridge-Public-Comment__Sam-Churchill_Oct-2019.pdf)
➤ Tunnels can off-load the congested I-5. Maybe 10%.
➤ Autonomous, rubber tired pods are here.
➤ They go the last mile and provide intercity links.
➤ Autonomous shuttles will connect Vancouver to Portland.
➤ They’re cheaper, faster and more flexible than buses.
➤ No new infrastructure. Rubber tires.
➤ Tesla’s “loop” may make commuting fast and convenient.
➤ Self driving vehicles will arrive before a new bridge. Guaranteed.
A bridge and highway system – built for the Model T – is not a solution.
- Shanghai Maglev
- Shanghai to Beijing by high-speed train
- Miami Maglev
- Hyperloop Explained
- Cold Fusion: Hyperloop
- Dallas-Ft Worth HyperLoop
- Chicago Loop to O’Hara
- Boring Company Video
- California HSR
- Oakland BART Tram (no driver)
- Siemens Sacramento Train Factory
- Transbay Tunnel
- Alaska Way Tunnel
- How Seattle chose the Bertha tunnel alternative
- Seattle Monorail
- Vancouver BC’s driverless SkyTrain
- Building the Chunnel
- A Ride on the Heathrow Pods
- London Heathrow ULTra Pods
- Autonomous NAVYA Shuttle
- Las Vegas launches NAVYA
- China vs USA Autonomous Vehicles
- How highways wrecked American cities