ACES, Driverless Vans Fatal Flaws

The Feb 4^th Seattle Times editorial page submission “Let technology take the wheel for safer, less-congested roads” includes a mirror view depiction of a woman with her feet on a car dashboard with the other seats empty, enjoying commuting in an “autonomous” car; hardly a recipe for reducing congestion.  It concludes people will “adopt transportation as a shared device to be summoned on demand” reducing the number of individually owned vehicles”.  However it’s not clear how “autonomous” vehicles “summoned on demand” will reduce congestion.

Since there is very little actual data,  it’s also not clear how “experts” arrived at the conclusion “once 20% of the cars are “autonomous,” crashes will decrease by 90%, and traffic will improve with fewer traffic snarls”.    The editorial claims the area should lead in implementing “autonomous, connected, electric, and shared” vehicles (ACES) as a way to reduce congestion.   Surely more data is needed to assure  “cars without drivers” would make a significant difference before making any significant ACES commitments.

Former King County Metro Transit director, Chuck Collins, is even more optimistic about “driverless, electric vans”.  His Feb 5^th editorial suggests,  “A future free of congestion thanks to driverless vans can be ours”.  While initially claiming, ”It is hard to say precisely what it would take to free up our streets and highways,” proposes a solution; “25,000 van pools to carry 500,000 passengers round trip, or 1 million (passenger) trips a day”.  Each van would make 10 one-way trips daily, carrying 5 passengers an average of 11 miles..   He neglects to mention whether or how these “trips” would respond to “summoned on demand” requests proposed above.    

His “van-trip” approach for commuters in the morning would require someone or something identify either a route to or the location of 5 commuters living within reasonable distance of each other, all wanting to go at the same time, to destinations reasonable close to each other.   Similarly, the afternoon van “trip” would also require 5 commuters within reasonable proximity whose departure time and destination are also compatible.  Collins’s assumption each van could make 5 such trips during the morning and afternoon commutes seems rather “optimistic”.

It’s even more “optimistic” to claim 25,000 vans could carry 500,000 commuters each morning and afternoon.  Doing so would require a huge database of potential van commuters.  Identifying 200,000 routes, each of which need to provide acceptable commutes to and from desired destinations for 5 passengers.   His comment, “In the last 40 years, employment has profoundly decentralized to new employment centers: Bothell, Overlake, the Kent Valley, Factoria, South Kirkland, Tukwila, and countless other centers” makes scheduling 200,000 routes to the needed destinations even more “difficult”.    

Collins also expresses concerns  that “Most (commuters)  will run a gauntlet through the congestion of Interstates 5 and 405, Highway 167 and many other stretches of bogged-down traffic”.  (Its not clear how that concern is consistent with his claim 93% of commuters are heading someplace other than downtown Seattle.) However, driver-less vans are also not the best option for those concerns.

Reducing  roadway congestion nominally requires limiting the number of vehicles to around 2000 per hour.  While it’s not clear how autonomous cars would increase lane capacity, a 5-passenger van only results in 4 fewer vehicles.  By comparison, a 70-ft articulated bus can accommodate up to 119 sitting and riding commuters, increasing the lane capacity by nearly 30 times that of the van.  

The lane capacity increase from any reasonably sized bus dwarfs that of a van, whether driver-less or not.   Also, congestion at destinations in Seattle,  Bellevue and other major locations  would be far less with buses rather than vans.  Up to 900 buses an hour are routed into Manhattan on a single lane.    Clearly buses provide far more  of the public transit capacity needed to reduce congestion on the areas roadways.  (They’re also far better than light rail as attested to by Collin’s estimate for 30,000 daily riders with ST3 extensions.)  The reduced congestion with bus routes to Seattle, Bellevue, and other major destinations will also benefit commuting to other “decentralized employment centers”.

Their only limitation on ridership is the need to provide commuters with access.  Commuters throughout the area need to be surveyed to identify how best to provide access.  Rather than attempting to identify 200,000 routes, the results could be used to prioritize locations for 1000-stall pay-to-park lots and bus stops within walking distance for large numbers of commuters. 

Those paying for parking would have a reserved parking stall and priority access to a bus.   Those using the bus stop could buy monthly or yearly “Assured Access” fares to also guarantee access to their preferred route.   In both cases the funds from those paying for parking or assured access will cover operating costs allowing many others to ride free.   A sure recipe for the increased ridership needed to reduce congestion.

The bottom line is reducing congestion is not that complicated.   Either add additional lanes or increase capacity of existing lanes.  ACES does neither and any increase in lane capacity with driverless 5-passenger vans will be dwarfed by increases from added buses.  While driver-less vans may have some use in Uber or Lyft types of operation,  they’re “unlikely” to significantly reduce congestion on the area’s roadways. 

Implementing the improved bus service requires converting Sound Transit from one that manages the construction of light rail extensions to one that increases public transit capacity with added parking and bus routes.   The most viable way to do so is to require an audit that would expose the total failure of their light rail spine to reduce congestion.