Dedicated I-5 lanes and a midroute vehicle swap could move riders between Los Angeles and San Francisco years before the planned train is available.
This post was also published in Inc.
Much has been written about the travails of California high-speed rail between San Francisco
and Los Angeles, with an initial projected cost of $33 billion and estimated completion by
2020. That original
estimate from 2008 has now expanded to a projected cost of up to $231 billion and an estimated start of operations in 2040.
What if there’s a better way to achieve high-speed transport with technology that’s already
been deployed?
Transportation and AI technology have accelerated in the eighteen years since the 2008
California Proposition that approved the high-speed rail project. In the last six months, both
Tesla Robotaxis and Waymo self-driving taxis have started operating on highways. Combined
with innovations in battery range, self-driving vehicles now present a viable public
transportation alternative to the large capital expense and long timelines of high-speed rail.
I’m a serial entrepreneur and tech investor. Here’s what I think a car-powered replacement for
high-speed rail could look like.
Use dedicated lanes on existing highways for high-speed travel
There is already a multi-lane, high capacity freeway between San Francisco and Los Angeles:
the I-5. Transportation departments have a lot of experience isolating left lanes into separate
high occupancy or paid express lanes. It’s just as easy to separate the two left lanes of the I-5
into dedicated lanes available only to self-driving vehicles that can safely operate at high
speeds.
The proposed California high-speed train would complete the Los Angeles to San Francisco
trip in 2 hours and 45 minutes, but getting to and from the train station at each end is likely
30 minutes on each side, and getting to the station 15-30 minutes early in order to catch the
train, for a total of 4 hours and 15 minutes. A self-driving vehicle making the same trip would
take about 6 hours.
This is indeed up to two hours more in total travel time using present technology, but
consider the fact that high-speed rail will likely not be available for another fourteen years. In that time, batteries will continue to improve, safety features will likely improve, and selfdriving vehicles should be able to make up the time by traveling even faster on those I-5
straightaways without using up the whole battery.
Another plus side is that people can set their own schedule for when they want to travel
rather than relying on a train schedule, which can be a big timesaver for business travelers
Create a large transfer point to split the journey across two fully-charged vehicles
Self-driving electric vehicles like long-range Teslas can only make it half-way between and San
Francisco and LA before having to recharge. Rather than expecting passengers to wait an
hour to charge, a large transfer facility can be established where passengers can swap into a
new fully charged vehicle.
Passengers would simply get out of the vehicle they are in, use the facilities, get a drink or
snack, and then simply get into a different fully charged vehicle for the rest of the journey.
The first vehicle can then get cleaned and start charging for a subsequent transfer passenger.
The cost and sustainability argument
Building out a new rail line and stations is extremely resource intensive and consumes vast
quantities of concrete and steel, and vast quantities of energy and natural resources.
Though cost-analysis studies are needed, I hypothesize that using existing electric vehicles
and charging infrastructure (which also has a sustainability cost) would be much simpler and
less resource intensive. The vehicles can be used for regular robotaxi services during non-
peak times and also be sold into an after market after some battery decay, for maximal
resource usage and reusability.
And self-driving robotaxis are already at a comparable cost to high-speed rail.
The estimated cost for California high speed rail is $231 billion and likely to keep going up,
and the estimated cost to ride is $81-$150 to travel 494 miles by high speed rail. Realistically,
the Boston to DC Acela train is comparable route at 457 miles and charges $100-$645. Tesla
Robo Taxi costs $3.00 base fare + $1.40 per mile, so San Francisco to Los Angeles would be
$695 at current prices.
The cost of building high speed rail keeps going up, while the cost of self driving vehicles
keeps going down. Consider that the $231 billion allocated to build out California high-speed
rail could instead pay for 332 million robotaxi rides at the current price, or 10-15 years of the
estimated future ridership of the high speed rail line.
It’s available now!
The current projection for high speed rail between Los Angeles and San Francisco is projected
over ten years from now in 2040 and will likely continue to be delayed. Electric vehicles, full selfdriving, rapid charging, and isolated freeway lanes are available today and could easily be
deployed in less than a year. The batteries, speed, and range of self-driving electric vehicles
will continue to improve every year and within ten years will certainly start to equal the total
transit time of high-speed rail well before high-speed rail would even be available in 2040.