Refueling Automated and

Autonomous Trucks - Challenges

and Opportunities
By: Michael D’Orazio, Cornelius Bittersohl, Max Rasovsky, Raul Kraus
White päper

Table of contents

1. The Diesel vs. Electric Timeline 4

2. Automated/Autonomous Truck
Development Timeline 6

3. Taking a Look at Refueling 7

4. Refueling Attendants and Machines 10

5. New Needs & Opportunities 11

6. Looking Ahead to Electric 12

7. Setting Up for Success 13

2
White päper

As multiple companies continue to make progress on developing and testing autonomous

trucks, many challenges remain to be overcome, and not all of them are strictly technology
limitations. Operating heavy duty tractor-trailers involves far more than simply driving from
point A to point B. A range of logistical considerations also need to be taken into account
as autonomy is on track to take the driver out the cabin, from loading and inspection to
weigh station and cargo sign-off. Among these considerations, an often overlooked one is
refueling. In this study, we take an in-depth look at how refueling will need to be planned
for and handled for autonomous trucks to be successful:

Gas station operators will need to find new solutions to refueling of automated trucks,
including utilizing attendants, automating the task, automating payments, and finding
optimal locations to maximize autonomous truck traffic and minimize operating costs.

Truck fleet operators will need to consider different factors to maximize fleet
effectiveness and return on investment including new business models, partnerships with
fuel stations, and optimization of routes.

Truck manufacturers will also have to be mindful of design challenges and new
market demands that the altered use of trucks through autonomy will lead to.

3
White päper

1. The Diesel vs. Electric Timeline

There has been a large amount of press in the last year surrounding the development of
electric heavy-duty trucks, with prototypes from Freightliner (Daimler Trucks), Volvo, and
Tesla making the rounds at conferences across the US. However, based on realistic
development schedules, setbacks, and current readiness, we expect customer-ready
electric truck development to follow a relatively slow roll-out timeline.

umlaut industry projection through 2030

Additionally, electric trucks will be impacted by the same range and weight concerns that
slowed adoption of electric passenger vehicles for years. Modern Class 8 trucks with a basic
aerodynamic package can get up to 11 miles per gallon of fuel efficiency, but usually
average around 8 miles per gallon under realistic conditions. Carrying enough fuel for trips
of 1,000 miles or more at this efficiency level has become standard. Meanwhile, initial
estimates and tests of all-electric Class 8 truck range under normal hauling conditions are
around 250 miles for first-generation trucks.

A major component of this seemingly small range is the “rocket equation” equivalent for
large batteries. Just as rockets are limited by the fact that they have to lift the weight of all

4
White päper

the fuel they carry, electric vehicles are limited by the fact that they have to accelerate and
decelerate the weight of the batteries that power them. This means that the bigger the
battery used, the more weight the vehicle has to get moving and slow down, which lowers
overall vehicle efficiency. With current battery technology, first-generation electric trucks
with a 400kWh battery pack are expected to average around 2.1 miles/kWh of range
efficiency. The bigger the battery pack, the lower this number gets. It will take some time
for technology improvements to change this equation significantly. Based on battery cost
and power density improvements, range improvement for electric trucks is expected to
follow a slow upward trend over time.

umlaut industry projection through 2030

What this means in practical terms is that electric trucks will initially be used primarily for
short-haul trips in more urban environments and between ports and distribution centers.
Diesel trucks will still be favored for long-haul trucking on highways until battery range nears
parity with 125-gallon diesel tanks.

5
White päper

2. Automated/Autonomous Truck
Development Timeline
An interesting outcome of this analysis becomes clear when looking at the overlap with
expected autonomous truck development timelines and use cases. Autonomous
development for vehicles began in earnest in the 1980s, with each milestone advancement
being an order of magnitude more difficult than the one before. Based on current technical
development tracks and capabilities, we agree with other market assessments predicting
that autonomous truck development and roll-out will be based on design domains rather
than “all at once” approaches.

What this means in real-world settings is that trucks will first be fully autonomous-ready in
highway-only modes (“exit to exit”) best suited to long-haul routes rather than in more
complicated urban settings. This will likely happen in the mid-2020s, followed by additional
development and testing before semi-urban autonomous driving will be a reality for trucks
in the early 2030s.

An obvious result of these overlapping timelines is that diesel heavy duty trucks will likely
lead the way in long-haul autonomous driving for some time until electric infrastructure and
batteries catch up.

6
White päper

3. Taking a Look at Refueling

Drivers of passenger cars are used to putting gas in the tank when the low fuel indicator
comes on, but in the world of long-haul trucking, refueling is a task that often needs to be
planned and optimized. A Class 8 truck driver can easily be filling up over 100 gallons of
diesel at a time, sometimes in multiple tanks, using the stop time not only to get more fuel
but also to eat, use the bathroom, take a shower, exercise, and sometimes sleep. True truck
stops are small service economies unto themselves where adding diesel to the tank is only a
small part of the overall transaction. But autonomous trucks will change this model entirely.

Autonomous trucks
won’t need traditional
human-oriented
services like food and
bathrooms, shifting
truck stop business
models away from
higher-margin goods to
specialty services like
sensor inspection and
cleaning. Additionally,
owners of autonomous
truck fleets will always
be looking to reduce their operating costs as much as possible. With heavy duty trucks, this
means not only reducing operating costs by not having to pay human drivers, but also
maximizing return on investment by reducing the amount of time trucks are idle. Unlike
human drivers who are bound by strict rules governing maximum shift and total time behind
the wheel, autonomous trucks will be able to drive the highways of America 24/7 and only
stop to switch loads, repair, and refuel. Based on current trucking economics, we estimate
that driverless trucks can run with 30-40% lower operating costs than their human-driven
counterparts. To maximize this cost reduction, reducing fuel stop time as much as possible
is the name of the game.

Compounding this new reality will be a marked rise in the percentage of goods transport
miles given to trucks rather than rail and water transport. To see why, we need to look at
the economics of freight transport and the key criteria that drive selection of one transport
method over another.

7
White päper

1. Base Cost

Average costs per ton-mile of transport varies by load type, conditions, and year, but
roughly break out as follows1:

Air freight: $125

Truck cargo: $18

Rail cargo: $4

Domestic Waterways: $3

Cargo transport via trucks today is more than four times as expensive as transport via rail.
However, this is largely due to labor costs, which often account for upwards of 40% of per
ton-mile cost for trucks. At the same time, trucks are over a hundred dollars per ton-mile
cheaper than air transport.
2. Transport Speed

Air is obviously the fastest way to move goods long distances, but how much faster is it in
reality? It really depends on how far your goods need to move and processing times at air
hubs. In many cases, freight can be moved several hundred miles by truck in the same time
it would take to get it to and from an aircraft. We estimate that current freight break-even
transit time with air occurs around 400 miles, with 1-day delays for each subsequent 500
miles traveled. However, this delay is at least partially due to truck driver shift limits. For a
truck that can drive around the clock, the equation shifts to 1-day delays for each
subsequent 1,200 miles traveled. Rail and water transport are obviously far slower due to
both lower travel speeds and the need to wait for cargo containers to be bundled into full
loads prior to departure and then unbundled at depots/ports. It is worth noting that 69%
of freight transport by goods value occurs within 500 miles of the shipping origin2, which is
perhaps the leading reason for the continued dominance of trucks as the go-to shipping
solution in the US.

3. Scheduling Flexibility

The final factor is schedule flexibility, which can be a huge concern for transporting time-
sensitive freight. Transporting freight via rail and water requires containerizing goods at rail

1
https://bts.gov/content/average-freight-revenue-ton-mile
2
FHWA Analysis:
https://ops.fhwa.dot.gov/freight/freight_analysis/nat_freight_stats/docs/13factsfigures/pdfs/fff2013_highres.p
df

8
White päper

depots and ports, then waiting for available locomotives or ships to be completely loaded
and unloaded. Additionally, trains and cargo ships often make multiple stops along a
freight route to drop off and pick up cargo, widening delivery windows by several days.
Freight via trucks, on the other hand, can easily be point-to-point and trucks can depart
on the exact schedule needed by shippers.

With truck-based freight transport already being favored over rail and sea for its quicker
transport times and more flexible scheduling options, the key factor preventing more freight
from being delivered via trucks is cost. But autonomous trucks stand to significantly reduce
truck-based freight transport cost over time due to the elimination of driver pay, lower
accident rates, and more efficient use of trucks themselves. As these cost reducing factors
increasingly come into play along with transport speed improvements due to round-the-
clock driving, transporting freight via trucks will become increasingly attractive and take
market share away from competing transport methods. We expect that autonomous trucks
will account for a significant increase in percentage of freight miles filled by trucks over the
next 15 years.

umlaut 20-year industry projection

This, combined with the previously mentioned push for “fast refueling only” services at fuel
stations will drive the development of a new generation of truck stops that cater specifically
to the needs of autonomous trucks. These new stations will differ from current ones in
several key ways.

9
White päper

4. Refueling Attendants and

Machines
Outside of a few states in the United States where pumping your own gas is illegal, gas
station attendants are a rare sight today. However, autonomous trucks will likely
necessitate a return to the station attendant model until practical robotics catch up with
cost effectiveness and demand. Human attendants will need to be on-duty around the
clock to refuel autonomous trucks as they pull in, manage payment by the truck operator,
and perform additional services like spot inspections as needed. Based on analysis of labor
in current markets where gas station attendants manage pumping and payment, we expect
that strategically placed truck stops will require an average of 1 attendant for every 5
pumps in active operation.

We also anticipate that station

attendants will return to the true “full
service” model of previous decades,
where basic maintenance operations
like tire pressure checking and
optimization, fluid level checking,
and sensor/lens cleaning become
add-on services at a small cost while
trucks are refueling. Without a
human driver to perform these tasks,
fuel stops become a logical point to
bring humans into the maintenance loop without impacting overall on-the-road time and
efficiency. This model will likely continue even with electric autonomous trucks due to
similar maintenance requirements and even longer refuel/recharge times per stop.

Over time, robotic pumps (whether for fuel or electricity) will take over the role of human
attendants for autonomous trucks. Tesla, for example, developed a proof of concept
automated plug using an articulated cable, similar to a tentacle, that can automatically
seek and connect to a charge port within its range. However, the cost, complexity, and
failure modes of devices like this will likely prevent them from gaining significant traction
over relatively cheap human attendant labor for some time. Based on expected labor costs
and remaining challenges for automated pumps, we expect that human attendants will be
the go-to model for autonomous truck stops until the 2030s, at which point economics will
naturally start promoting reduction of human labor.

10
White päper

5. New Needs & Opportunities

In addition to attendant-based services for autonomous trucks, fuel stops targeting this
new market will need to consider several other factors to guarantee financial success and
future readiness. Chief among these considerations is location, which needs to be carefully
determined on multiple factors, including route usage, proximity to freight destinations,
labor costs, fuel costs, and autonomous-friendly access. This unique combination of needs
means that autonomous-oriented fuel stops will be relatively sparse in comparison to
existing fuel stops and need to be carefully selected for optimal placement.

On top of geographic
location and due to their
specialized nature,
autonomous truck fuel stops
will also need to be
constructed as larger
operations than a traditional
fuel stop. Operators will
need to consider
simultaneous refueling of
many trucks with a much
faster turnover rate than
human-driven trucks. Additionally, early autonomous trucks will need to make use of
guides, markings, and sensor indicators to effectively navigate to available pumps and exit
the stations effectively. For this reason, we believe autonomous-oriented truck stops will
need to be more akin to a strategically positioned fuel depot than a fuel stop with a few
pumps.

And moving beyond fuel stations, autonomous trucks themselves will need to change over
time. Manufacturers or retrofitters will need to develop sensors and software specifically
targeted at refueling scenarios in close proximity to fuel pumps and other trucks, including
identifying where to stop within reach of a hose, turning off the engine during refueling, and
knowing when it is safe to restart the engine and continue driving. Looking further ahead,
manufacturers and fuel station operators will also need to collaborate on new designs for
pumps or charging stations, hoses/cables, and receptables that allow for completely
autonomous refueling without a human attendant.

11
White päper

6. Looking Ahead to Electric

Even though we expect diesel to be the primary fuel source of long-haul autonomous
trucks for some time, electric trucks will eventually gain prominence in the market and
require recharging stations that can handle their unique needs. Autonomous truck stop
operators will need to take into account several considerations specifically for this new
breed of truck, including infrastructure needs and optimal locations.

While initial Class 8 truck

battery packs are being
developed at around a
400kWh capacity level, this will
increase over time to as much
as 1,000kWh or more as
battery technology improves.
Charging multiple trucks, each
accounting for a full megawatt-
hour of energy storage,
simultaneously is no small feat.
Electric infrastructure to achieve
this level of power transfer is
significant and may even require small power plants and grid-scale storage on-site or near
electric truck recharging stations to keep up with demand. Implementing this type of
infrastructure will require significant investment, whether operators want to develop new
locations or extend existing ones.

Additionally, electricity prices are far more volatile than diesel prices, both by geographic
location and time of day. Optimizing location, power usage vs. generation, and storage will
mean the difference between strong profitability and negative margins for operators of
electric truck recharging stations. New power management systems and business models
will need to be developed to take advantage of the new opportunities posed by this
market.

12
White päper

7. Setting Up for Success

Our team of experts is heavily involved in the development, planning, and strategic
implications of heavy duty electric and autonomous trucks. We have developed multiple
models for optimizing placement, infrastructure, and implementation of autonomous-ready
fuel stops over a spectrum of future scenarios. We can help clients prepare to take
advantage of this market, including:

• Finding optimal locations to maximize autonomous truck traffic and minimize

operating costs
• Determining ideal land area and physical layouts to prepare for autonomous truck
customers
• Developing strong business models to generate sustainable profits as the
autonomous truck market grows over the next decade
• Optimizing station employee and skills mixes and creating new services targeted
specifically at autonomous truck fleets

Picture credits

https://www.pexels.com/photo/lights-night-dark-gas-92077/
https://www.pexels.com/photo/7-eleven-store-photo-during-sunset-3116970/
https://www.pexels.com/photo/car-refill-transportation-gas-9796/
https://www.pexels.com/photo/person-pointing-on-white-textile-34753/
https://p2.piqsels.com/preview/512/288/64/electric-car-load-plug-in-connection-e-car-e-mobile.jpg

13
White päper

umlaut Inc.
25650 W Eleven Mile Rd
Suite 300
Southfield, MI 48034

[email protected] +1 404 626 2354

[email protected] +1 248 765 8380
[email protected] +1 248 794 2694
[email protected] +1 313 400 0912

www.umlaut.com

14