You are on page 1of 6

1.

Executive Summary

The primary goal of this project was to estimate the value of autonomy in market
segments that utilize low-speed vehicles in private spaces, while also identifying
promising use cases per segment. The Duke MEM CPP team identified and
explored 8 promising sectors: 7 market segments (airports, seaports, universities,
construction, distribution centers, nuclear facilities, and luxury hotels and resorts)
and 1 vehicular segment (street sweepers). In addition to standard secondary
research using web and library resources, primary research included reaching out to
~200 operations managers across these market segments and interviewing ~30.
Through this work, 5 of the aforementioned areas appeared to have the strongest
use cases and market potential for autonomy: airports, seaports, universities,
construction, and street sweepers. An estimate of the maximum number of vehicles
(as well as those best suited for autonomy) was developed for each segment.
Separately, the value of autonomy was researched via driver-wage and accident-
cost analyses. By coupling these analyses, the value of autonomy in each market
segment (again, the maximum value as well as the value for the best-suited use
cases) was developed. Below is a graphic showing the core estimates for the 4
market segments (where street sweepers, also a smaller segment, was excluded
from the summary below as being a vehicle type rather than a market segment):

All Vehicles Within Segment Vehicles Best Suited for Autonomy

Quantity - All Vehicles Quantity - Best Suited for Autonomy


(# units) (# units)
Airports Seaports Universities Construction Total % Airports Seaports Universities Construction Total %
Shuttle Bus 8,781 10,461 147,245 0 166,488 30% Shuttle Bus 8,781 10,461 147,245 0 166,488 68%
Vehicle Shuttle Van 6,720 0 36,782 0 43,502 8% Shuttle Van 6,720 0 36,782 0 43,502 18%
Quantity Trucks 9,600 9,206 0 0 18,806 3% Trucks 0 0 0 0 0 0%
Custom GSE 120,958 0 0 0 120,958 22% Custom GSE 0 0 0 0 0 0%
Car/SUV 22,080 94,150 0 0 116,230 21% Car/SUV 0 9,415 0 0 9,415 4%
Golf Carts 0 0 55,202 0 55,202 10% Golf Carts 0 0 0 0 0 0%
Forklifts 0 9,206 0 0 9,206 2% Forklifts 0 9,206 0 0 9,206 4%
Custom Trucks 0 0 0 15,760 15,760 3% Custom Trucks 0 0 0 15,760 15,760 6%
Total 168,138 123,023 239,230 15,760 546,151 100% Total 15,501 29,082 184,027 15,760 244,371 100%

Autonomy Value - All Vehicles Autonomy Value - Best-Suited Vehicles


($ millions) ($ millions)
Airports Seaports Universities Construction Total % Airports Seaports Universities Construction Total %
Shuttle Bus $650 $105 $1,637 $0 $2,392 14% Shuttle Bus $650 $105 $1,637 $0 $2,392 64%
Autonomy Shuttle Van $498 $0 $409 $0 $907 5% Shuttle Van $498 $0 $409 $0 $907 24%
Trucks $711 $92 $0 $0 $803 5% Trucks $0 $0 $0 $0 $0 0%
Value Custom GSE $8,955 $0 $0 $0 $8,955 54% Custom GSE $0 $0 $0 $0 $0 0%
Car/SUV $1,635 $942 $0 $0 $2,576 16% Car/SUV $0 $94 $0 $0 $94 3%
Golf Carts $0 $0 $614 $0 $614 4% Golf Carts $0 $0 $0 $0 $0 0%
Forklifts $0 $92 $0 $0 $92 1% Forklifts $0 $92 $0 $0 $92 2%
Custom Trucks $0 $0 $0 $263 $263 2% Custom Trucks $0 $0 $0 $263 $263 7%
Total $12,448 $1,230 $2,660 $263 $16,602 100% Total $1,148 $291 $2,046 $263 $3,748 100%

The leftmost tables (in blue) above show that nearly 550,000 vehicles exist within the
market segments analyzed, which would lead to a nearly $17 billion theoretical
maximum autonomy value. The rightmost tables (in brown) above show that nearly
250,000 vehicles exist within the market segments analyzed that are best suited for
autonomy, which would lead to a nearly $4 billion more near-term autonomy value.

The same data can be viewed in more graphical form as shown below.

4
All Vehicles Within Segment Vehicles Best Suited for Autonomy

Quantity - All Vehicles (units) Quantity - Best Suited for Autonomy (units)
180,000 180,000
160,000 160,000
140,000 140,000
120,000 120,000
100,000 100,000

Vehicle 80,000 80,000


60,000
60,000
Quantity 40,000 40,000
20,000 20,000
0 0
Shuttle Shuttle Trucks Custom Car/SUV Golf Carts Forklifts Custom Shuttle Shuttle Trucks Custom Car/SUV Golf Carts Forklifts Custom
Bus Van GSE Trucks Bus Van GSE Trucks

Airports Seaports Universities Construction Airports Seaports Universities Construction

Autonomy Value - All Vehicles ($ M) Autonomy Value - Best Suited Vehicles ($ M)


$3,000
$10,000
$9,000 $2,500
Autonomy $8,000
$7,000 $2,000
Value $6,000
$1,500
$5,000
$4,000 $1,000
$3,000
$2,000 $500
$1,000
$0
$0
Shuttle Shuttle Trucks Custom Car/SUV Golf Carts Forklifts Custom
Shuttle Shuttle Trucks Custom Car/SUV Golf Carts Forklifts Custom
Bus Van GSE Trucks
Bus Van GSE Trucks

Airports Seaports Universities Construction


Airports Seaports Universities Construction 8

Viewed in this form, one can make various inferences such as shuttle buses being a
key vehicle type and that universities and airports represent key segments for
autonomy value extraction. Custom ground support vehicles (GSEs) within the
airport segment will appear significant at first (and will perhaps be relevant in the
future), but are not considered the best-short-term value targets.

The data above are maximum market values for each case, which require market
penetration estimates to estimate how autonomy might be adopted over time. To
estimate such penetration, the team employed a Bass model to forecast how
autonomy might diffuse into the market over time as shown in the plot below.
Using Bass Model to Estimate Adoption of Autonomy Over Time
Innovation coef. 0.01 Massiani, 2015 (EV adopt)
Imitation coef. 0.1 Massiani, 2015 (EV adopt) Value of Autonomy in Cyngn Target Markets
CAGR vehicle market 3% Vehicle sector estimate $16.0

Year from Start Actual Year Adoption Value of Value of $14.0


Fraction Autonomy Autonomy
$12.0
Adopted Adopted
(theo max) - (most suited) - $10.0
in $B in $B
$8.0
1 2018 10% $1.6 $0.4
2 2019 18% $3.2 $0.7 $6.0
3 2020 26% $4.8 $1.1
$4.0
4 2021 33% $6.2 $1.4
5 2022 40% $7.7 $1.7 $2.0
6 2023 46% $9.1 $2.1
$0.0
7 2024 51% $10.5 $2.4 1 2 3 4 5 6 7 8 9 10
8 2025 56% $11.8 $2.7
Value of Autonomy Adopted (theo max) - in $B
9 2026 61% $13.1 $3.0
Value of Autonomy Adopted (most suited) - in $B
10 2027 65% $14.4 $3.3

Using conservative diffusion coefficients found in the literature related to electric


vehicles (a related technology), the adoption fraction column was computed. A
conservative compound annual growth rate (CAGR) was incorporated into the model
(since vehicle quantities will naturally grow over time). Finally, the previous
estimates were combined with this Bass-model adoption model to generate an
autonomy value estimate per year for 10 years. Note that the blue curve again
represents the maximum case (all vehicles) while the brown curve represents the
best-suited case (vehicles believed to convert to autonomy first). For example, the

5
maximum case would have a $1.6 billion value in 2018, growing to $14.4 billion by
2027. Similarly, the best-suited case would have a $400 million value in 2018,
growing to $3.3 billion by 2017.

All of these estimates (e.g., numbers of vehicles, autonomy value per vehicle,
adoption per year) are subject to modification and improvements as CYNGN further
researches this topic and varies modeling assumptions. Thus, an Excel-based
workbook is provided that can be use to facilitate such modeling moving forward.

Leading takeways from this research include the following:


• Airports and universities represent the best long-term and short-term
autonomy value, respectively. Key contacts made during this research should
aid further market estimations and partnership explorations in these and other
sectors.
• Shuttle buses represent the best autonomy opportunity by vehicle type, while
street sweepers are also of interest. Reasonably low market fragmentation
(at least for some of the top manufacturers) would suggest that a few
partnership calls within these sectors could provide high business-
development value.
• Additional work that CYNGN can do to further enhance market estimates and
business development include the following:
o adding benefits and increased efficiency to the autonomy model
o continued competitive analysis (especially on firms like Navya)
o further data mining of market reports already identified including the
LSLEV report
o updating/refining use case and vehicle quantity estimates per sector
o considering if accident value varies by industry (beyond just the airline
industry)
o exploring the airport GSE sector more deeply for promising use cases

Other topics covered in the report include the low speed light electric vehicle
(LSLEV) market (identification of a report that can be of future value), competitive
analysis, and regulatory research. The final deliverables package includes this
report, a companion presentation, and a Google Drive folder with the Excel model
and associated research findings. See the appendices within this report for
additional details as well, including a list of Google Drive contents.

The Duke MEM CPP Team appreciated the opportunity to work with CYNGN during
this Fall 2017 project.

6
2. Motivation & Objectives

Cyngn seeks to provide autonomous software and sensor solutions to segments of


the driverless vehicle market with a more near-term adoption horizon (i.e., areas that
may have less stringent regulations and risks than, for example, public highways).
Thus, a key goal of this project was to estimate the value of autonomy in market
segments that utilize low-speed vehicles in private spaces.

The scope of the project was defined in the initial call between the CYNGN Team and
Duke CPP team. The team agreed to focus primarily on the following key aspects:
• Total addressable market for autonomy in promising private spaces
o Identify promising market segments
o Estimate vehicles per segment
o Estimate autonomy value per vehicle
o Estimate autonomy value per segment
• Competitive analysis
• Pricing Analysis
Our focus was to identify promising market segments for CYNGN, identify the different
vehicle types and use cases in these segments, calculate vehicle estimates – overall
and by region – and combine these with a value of autonomy calculations to derive
numbers for the total addressable market. We also further divided this to identify a
maximum market for autonomous vehicles with target spaces overall across all vehicle
types and the best near-term autonomous use case.

While working on this project, we had to navigate through challenges. We had to steer
away from “popular” literature which usually presented information that focused on
autonomous solutions for passenger cars (e.g., Google, Lyft, Uber, & car companies).
We also had to maintain CYNGN’s anonymity while conducting primary research in
the promising market segments.

3. Research & Analysis Strategy

Our Research and Analysis Strategy included the following key steps:
1. Identifying promising vertical markets that meet Cyngn’s Specification
2. Conduct secondary and primary research per segment
• Estimate vehicle types and quantities per segment
• Find way to extrapolate estimates to find vehicle TAM for segment
3. Research topics to estimate value of autonomy
• Savings to fleet owner by eliminating driver wages/benefits
• Savings to fleet owner by decreasing risk of accidents
4. Develop full TAM model for vehicle autonomy (including all segments)
5. Research and summarize data on other topics including competitive analysis,
etc.
In order to find promising market segments, we started with finding applications
relevant to CYNGN as shown in the figure below.

7
Figure 1: Finding Relevant Applications

After careful consideration, we selected the following vertical markets for further
analysis.
• Airports
• Seaports
• University Campuses
• Luxury Hotels & Resorts
• Nuclear Facilities
• Distribution Centers
• Construction
We selected these market segments as they fit the criteria of having low-speed
vehicles in private spaces. We anticipated the adoption of autonomy in these
segments and also noted competitor activities to provide autonomous solutions within
these segments.

The team identified around 200 operations managers across these segments, and
conducted interviews with ~30. We conducted surveys with selected segments and
logistics companies to get more information about these industries. Our main focus
during interviews and research were the following:
1. Estimate vehicle types & quantities
2. Identify promising use cases for autonomy
To estimate vehicles and market size per market segment, we followed the steps as
shown in Table 1 below.

8
# STEP DESCRIPTION EXAMPLE FOR “AIRPORT”
MARKET SEGMENT (NUMBERS
FOR EXAMPLE ONLY)

1 Identify Top Sites Find (or develop) list of top sites Top 50 List of Global Airports by
(within some geography) by some Passenger Traffic
activity metric.

2 Obtain Vehicle Conduct secondary and primary Identified 9 vehicle types, and
Quantity by Type research to determine types of obtained quantity estimates per type
(VQxT) Data for vehicles used and quantities for for Shanghai and San Francisco
Sample Sites multiple sites. Airports

3 Scale VQxT Data With activity metric as a scaling Develop VQxT table for San
Across Top Sites factor (and a VQxT estimate for Francisco airport and scale across
an average or estimated case), top 50 airports using passenger traffic
calculate VQxT values for all top as scaling factor
sites.

4 Scale VQxT Data for Determine share of global market Top 50 Global Airports represent 35%
Global Estimate the top sites command, and scale of global passenger traffic, allowing
for global estimate global VQxT estimates

5 Estimate Autonomy Determine how much eliminating Shuttle bus drivers cost
Value Per Vehicle driver wages and decreasing risk $26,000/year/vehicle. Accident risk
(wage & accident of accidents will save the fleet equates to $5,000/year/vehicle. Total
savings) owner is $30,000/year/vehicle.

6 Estimate Max Global Multiply autonomy value per 170,000 global airport vehicles x
Autonomy Value for vehicle by number of all vehicles $30,000/year/vehicle autonomy
Segment worldwide within the segment savings = $5.1 B

7 Estimate Total Exclude from Max Global 15,,000 global airport shuttle buses x
Addressable Global Autonomy estimate those $30,000/year/vehicle autonomy
Autonomy Value for vehicles that seem unlikely to savings = $450 M
Segment adopt autonomy

8 Estimate % Adoption Select proper innovation (p) and If using p=0.01 and q=0.1 (Gross
per Year Using Bass imitation (q) coefficients to apply 2008 for EVs):
Model Bass Model to estimate fraction of
market that will adopt autonomy Year 1 = 10%, Year 2 = 16%, Year 3
over time = 26%, Year 4 = 33%, Year 5 = 40%,
Year 6 = 46%, Year 7 = 51%, Year 8
= 56%, Year 9 = 61%, Year 10 = 65%

You might also like