Industry Where Applying Models of Technology Diffusion Michael Smitka Economics Washington and Lee University February 16 2017 Initial version presented 21 Nov 2016 at Collegio Carlo ID: 580231
Download Presentation The PPT/PDF document "Disruptors in the Auto" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Disruptors in the Auto Industry? Where?Applying Models of Technology Diffusion
Michael
Smitka
, Economics
Washington and Lee University
February 16,
2017
Initial version
presented 21 Nov 2016 at
Collegio
Carlo
Alberto, Torino Italy
Funding
from
Munk
School of Global Affairs, University of TorontoSlide2The common view
We are now witnessing Elon Musks slow-motion disruption of the global auto industry
Title of a
reddit
“Futurology” post
But if it’s “slow-motion” can it be “disruption”??Slide3
Three Disruptors(Battery) Electric Vehicles: TeslaNew technology allowing new entrants who will kick ass
Mobility 2.0: Uber, Lyft, Zipcar and
Autolib
’
Vehicles sitting parked 22 hours a day
Monetizing $3 trillion in idle assets
Autonomy: Google car:
Eliminate accidents and congestion
Will do away with personal carsSlide4
wrong, WRONG, WRONGStep one: disruption poorly specified
define
Step two: diffusion assumed rapid
quantify for automotive tech
Step three: substitutes ignored
delineate
Step four: technical challenges underappreciated
detail
Step five: so what?
Incumbents will remain principal players
Prisoners dilemma: invest today, reap no gains tomorrow
Public
policy: governments may be able to sway R&D locationSlide5
Step Five: So What? (Part I: Firm Level)Prisoner’s Dilemma
OEMs invest today, up to $1
bil
per annum incremental expenses
Everyone is in the game
No one will gain a competitive
advantage
Costs will
rise, won’t
lead to higher margins
No near-term revenue so
worse
from net present value perspective
None
hold
back
After all, everyone’s doing it
But parts manufacturers may fare
better
SUM: All play,
all
see profits fallSlide6
Step Five: So What? (Part II: Regional Level)
Where will R&D take place?
BEVs and Autonomy require 15-20 years of additional R&D
Auto
firms setting up Silicon Valley
listening posts
Silicon Valley is setting up Detroit
engineering centers
Connected
car research
facilities:
“
M-City”
in Michigan, similar programs
in
Ontario
Policy encourages agglomeration in traditional US-Canada auto alley
Where
will battery production locate?
consumer electronics base
initial
core
= China (#1), Korea, Japan
NOT EU, NAFTA
But bulky so cell production migrating to US
Not true
(at present)
for “fine chemicals” = most profitable segmentSlide7
Step Five: So What? – Dismal Science warningManufacturing productivity ↑↑
Despite higher output
Factory jobs ↓↓
Policies that focus on standard blue-collar jobs
will not workSlide8
Step II: DisruptionNeed objective criteria forcorporate strategic relevance
Wall Street investment relevance
Does not preclude consumers benefiting more
quickly
Will new entrants comprise 25% of an
industry?
For the auto industry,
2
platform cycles = 12-16
years
25% means 25 million units annual production
Will profits fall 25
%?
For Wall Street, time horizon is
AT MOST
8 yearsSlide9
Disruption reality checkHave incumbents changed their behavior?
Piston
manufacturers
continue to invest in R&D
for technologies that will not come to market until the mid-2020s
continue
to invest in new capacity
they
expect higher piston output in
2030!
Despite
Uber,
have
Yellow Cabs disappeared? Rental companies vanished?
Despite ADAS (advanced
driver
assistance systems)
have
insurance rates fallen?Slide10
S-shaped Diffusion ProcessTechnologies go through a long evolution from
Initial development
early
commercialization
Improved
commercial productions
widespread
diffusion
Seminal
work:
Zvi
Grilliches
(1957) on hybrid corn adoption
Neighbors
follow
lead of early adopters
hybrids improved, extension services spread know-how
Eventually 100% diffusion
BUT
took
decades
His result proves robust
across
technologies
Pace of diffusion function
of product
cycleSlide11
(1-
)
Slide12
Technology rollout in the automotive industryNew technologies must be built into actual cars
Cars are only redesigned once every 4
years
Toyota
Ann Arbor works on one model a
year
Major redesigns
= new platforms = every
other cycle
BEVs allow
different
weight
distributions
need new platform
Light
truck redesigns less frequent: 6-10 years for F-150Slide13
MOTOR VEHICLE TECHNOLOGY ROLLOUT SIMULATION
new
BEV share
BEV output
BEV
in vehicle stock
stock
BEV stock
2020
100
1%
1
0%
1000
1
2023
109
2%
2
1%
1073
6
2025
116
5%
6
1%
1127
14
2027
123
12%
15
3%
1187
35
2028
127
20%
25
5%
1219
57
2029
130
29%
38
7%
1252
91
2030
134
41%
55
11%
1286
139
2031
138
53%
73
15%
1322
201
2032
143
62%
89
20%
1358
274
2033
147
70%
102
25%
1397
354
2036
160
79%
127
40%
1519
608
2039
175
81%
141
52%
1655
852 Slide14
Sum: Supply-side storyNormal diffusion process means disruption unlikelyLong automotive product cycle and durable good nature accentuate
Battery electric vehicles won’t constitute half the fleet until 2040
Autonomous vehicles will be even slower – technology not yet robust, much less cost effective
Demand-side factors matter, tooSlide15
Step Three (A): Electric Vehicle Substitutesbelow: Chevy Bolt: Interior, GutsSlide16Slide17Slide18Slide19
Battery Electric VehiclesChevy Bolt battery pack 960 lbs
for 200 mile range
Chevy
Cruze
gas tank 120
lbs
for 400 mile range
Batteries
alone
23
% of Bolt costs
($6000?)
No value proposition for consumers
Absent $7500 tax credit in US, $8000 cash subsidy in China
BEVs remain “Compliance Cars” (ZEV credits, CAFE credits)
NO successful new entrant (Tesla bleeds money)
Batteries: Incumbent consumer
electronics
producers
Controls: incumbents
car electronics firms (
Delphi, Bosch,
Denso)Slide20
Step Three (B): New Mobility SubstitutesCore business
not
new
Uber, Lyft,
Autolib
’,
ZipCar
:
fleet management
Car
companies have tried
(
eg
Ford owned Hertz)
They
were bad at
it
goal OEM high volume high prices
They didn’t have to own fleets to sell to fleets
Leave it to Enterprise Car Rental, best in
industry!
No disruption here for car companiesDisruptive for
government-supported Yellow Cab monopoliesSlide21Slide22
Uber case studyCar sharing present since start industry: Not new
Cell
phone
app technology
easily
replicated
Many software
tools dynamic pricing,
cf
OEConnection
No barriers to entry, no
Economies of Scale
Uber provides branding
So does Yellow Cab!!Slide23
Uber: ConclusionCost disadvantage
relative to
taxis
Can’t pay lower wages
Better reliability = larger
fleet
running
few
er miles
Underpricing rivals can’t continue
No entry
barriers
/
EOScale
= can’t price more
Without lower prices market won’t expand
Fundamentally
unsustainable business modelSlide24
Step Three (C): AutonomyPremise: transportation
is a utilitarian service
So who will buy?
Uber??
Small market in
developed countries
Fundamental tension: Cars have
never
been
utilitarian
Commodities carry thin marginsSlide25Slide26
Step Three (C): AutonomySo far 40
year rollout
process of
passive and active safety
Engines
run by Engine Control Unit
(1980-)
eSteer
instead of hydraulic steering
(1999)
Electrically controlled braking
= ESC
Electronic Stability Control
Active cruise
control (2000): radar
from
GM Hughes Electronics
Elements first introduced in luxury vehicles
Migrated if/when they find consumer acceptance
Virtuous circle of higher volumes leading to lower costs and wider adoption
Core
features long in useSlide27
Autonomy: Technical HurdlesHuman interface challenges to ADAS (Advanced Driver Assist Systems) Driver distraction:
can’t retake control quickly
Impedes
stepwise rollout
of autonomy components
Diminishing returns as features added short of full autonomy
Full autonomy (SAE Levels IV-V) faces huge technological barriers
Requires combination of Lidar, Vision and Radar systems
CPU for algorithms, GPU for sensor data processing
Not robust to “unusual” situations (individuals directing traffic)
Not robust to poor weather, poor roads
Needs
geofenced
areasSlide28
Autonomy: incumbents dominateOpportunities for new entry
Sensors, processors: mainly incumbent suppliers
Intel, Freescale/NXP,
Renesas
,
NEW: NVDIA
Software providers: ditto – Delphi Bosch Continents
NEW: Google
, Apple
Why should car companies get directly involved????
Not
facing monopolist suppliersSlide29
Peak Auto: Prisoner’s Dilemma reduxBEVs and Autonomy require lots of money up frontIncremental
revenue in
distant
future
All are players = margins won’t improve
Meanwhile CAE
(
computer aided engineering
)
lowers
new model
costs
Flood
of models, quick response to new
trends reduces
margins
I see no way out
Peak auto
cash-rich car
companies do
stupid things In past peaks bought car rental firms, aerospace firms, dealerships, banksSlide30
Long-run Challenges: Value ChainUpstream: SuppliersCan they maintain margins?
Dealerships
: maybe within 10 year “disruption”
Rise of large dealership
groups
Meanwhile information revolution
margin
compression
Service disruptors
More complex vehicles require sophisticated
tools
While reliability means fewer repairs
Will eliminate mom-and-pop repair shopsSlide31
Public PolicyGlobal industryWas multinational, operations everywhere, integrated nowhere
Today global vehicle programs, global
suppliers
Will it be “Trumped”?
Regional
structure:
end
of “branch plant” structure
Coalescing around “auto alley” in Europe, NAFTA (China unclear)
R&D location
Consolidation with globalization
Detroit more important than any time in past 50 years
Can policy keep high-value-added jobs here
?