Embed Size (px)
Authors: Uma Ayyer, Sajid Husain, Tao Jiang, Srdjan Petrovic, Anand Tolani
Self-Driving Cars: Disruptive vs Incremental
This work was created in an open classroom environment as part of the Engineering Leadership Professional Program (ELPP) developed and led by Prof. Ikhlaq Sidhu at UC Berkeley. There should be no proprietary informaEon contained in this work. No informaEon contained in this work is intended to affect or influence public relaEons with any firm affiliated with any of the authors. The views represented are those of the authors alone and do not reflect those of the University of California Berkeley.
“Autonomous Drive is about relieving motorists of everyday tasks, par9cularly in congested or long-‐distance situa9ons. The driver remains in control, at the wheel, of a car that is capable of doing more things automa9cally. Self-‐driving cars, by comparison, don’t require any human interven9on – and remain a long-‐way from commercial reality. They are suitable only for 9ghtly-‐controlled road environments, at slow speeds, and face a regulatory minefield.”
-‐ Carlos Ghosn, CEO of Nissan Motor Co, Ltd.
INCREM
ENTAL VS D
ISRUPTIVE STRATEGY
Automakers Google
Cars that assist humans while driving. Cars that taxi humans around.
TECHNOLO
GY DIFFEREN
CES Automakers Google
● Pre-‐mapped world: observing deltas.
● High-‐precision lidar technology. ($$$)
● RealEme view of the world. ● Sonar, radar, and camera
technology.
BU
SIN
ES
S SU
MM
AR
Y/TE
CH
NIC
AL E
NV
IRO
NM
EN
T
The adjacent technology industries to watch in this space are the sensors manufacturers, big data, silicon vendors with compute power in the embedded space and companies working on artificial intelligence, Google is moving forward with its plan to change the game by providing autonomous mobility services rather than selling cars. It plans to release this technology within 4 years. Auto-manufacturers predict that fully autonomous vehicles may not reach mainstream for at least another decade. The current sensor systems are very expensive. They are working on incremental automation technologies but not a driverless future.
Business Summary Technology Investment Opportunities
Google’s Strategy Auto Manufacturer’s Strategy
Self-‐Driving Cars
Sensors
Big Data
Compute
AI & RoboEcs
• Provide autonomous mobility services such as “robo taxi”
• Producing and selling specialized maps and software
• Technology for monitoring systems to reduce congestion
• Technology for robotics (probabilistic inference, planning & search, localization, tracking & control)
• Provide incremental automation capabilities in stages but not fully driverless
• Adaptive cruise control systems using radar (similar to auto-pilot on airplanes)
• Lane change assist • Automatic parallel-parking using imaging
sensors • Vehicle-Vehicle Communication
AUTOMATED PARKING ASSIST (Now) Toyota, Ford, Mercedes-‐Benz, Volkswagen,
Tesla
AdapFve Cruise Control(2016) BMW, GM, Volkswagen, Mercedes-‐Benz, Toyota
Highway Assist (2018) Nissan, Mercedes-‐Benz, GM
Driverless Car (2030) Nissan, Mercedes-‐Benz
Highway Auto Driving (2020) Nissan, Mercedes-‐Benz
ROADM
AP Google Automakers
Driverless Car (Now to 2020)
Market DisrupFon due to SDC Market Summary
Total Car Market Size PredicFon
MA
RK
ET S
UM
MA
RY/A
NA
LYS
IS
0 10 20 30 40 50 60 70 80 90
100 110 120 130 140
2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 2036
TradiEonal Semi Autonoums SDC
• Semi Autonomous Features bring in $30B addiEonal revenue in 2014
• Semi Autonomous features Expected to grow to $250B by 2030
• Google SDC is in prototype stage in 2014 • SDC will bring in addiEonal $80B revenue
by 2030 • 75% of car by 2035 will be either Semi
Autonomous or SDC • New entrant Google will capture 8% of
total Car market by 2035
• $200B Auto Insurance industry will transformed as Premium will go down due to less accidents
• Rental Car Industry, Taxi Service and Ride share will merged and evolve into Robo-‐Taxi Model.
• Auto Service industry will be consolidated into few big Automated service companies
• SoVware/Electronics content of car will be 50% of the cost of car
Millions Unit
OpportuniFes for Supplier of Components for SDC
** Data from mulEple sources: IHS and Market research
RE
GU
LATOR
Y FA
CTO
RS: S
TATE AN
D FE
DE
RA
L Bill Year License Routes Driver’s Seat
Occupant(1) Distracted
Driving Ban(2) Vehicle OEM
Liability Safety & OperaFng Standards
NV AB 511 SB 140
✓ 2011 TesEng for now Designated state highways
Lic driver & passenger req.
exempt when autonomous
exempt TBD by NV DOT
CA SB 1298 ✓ 2012 Mandate for tesEng and public operaFon
Not explicitly restricted
Lic. driver req. exempt when autonomous
not exempt Due from CA DMV before Jan 2015(3)
FL HB 1207 SB 1768
2012 TesEng only Not explicitly restricted
Lic. driver req. exempt Draier report was due Feb 2014
MI SB 169 ✗ 2013 TesEng only Not explicitly restricted
Lic. driver req. MI DOT to submit report by Feb 2016
CO SB 13-‐016 ✗ 2013 TesEng only Lic. driver req. with kill switch
exempt when autonomous
Bill indefinitely tabled by Sen. Brophy ciEng transportaEon commijee opposiEon
(1) Driver’s seat occupant must be a licensed driver with access to steering, throale, and brake controls. (2) Ban on use of cell phones and texFng while driver is operaFng the vehicle. (3) Intent for public operaFon by Jun 2015.
State Legislation
• “Preliminary Statement of Policy” 14-13 released May 30th, 2013. • Begun 4 year safety standards study, 1st phase to be completed in 2017. • Research includes “vehicle to vehicle” communication technology. • Encourages states to allow testing of self-driving. • Does NOT recommend states currently allow public operation. • Else, licensed occupant in driver’s seat should be required. • Suggests operators have special training and licensing
Federal : NHTSA
Level DefiniFon Example / ExplanaFon
Level 0 No AutomaEon driver always in complete control
Level 1 FuncEon-‐specific AutomaEon e.g. stability control or brake assist
Level 2 Combined FuncEon AutomaEon two or more automated funcEons
Level 3 Limited Self-‐Driving AutomaEon “the Google car” 2013, BMW X5
Level 4 Full Self-‐Driving AutomaEon driver not available at any Eme hjp://cyberlaw.stanford.edu/wiki/index.php/Automated_Driving:_LegislaEve_and_Regulatory_AcEon
Federal : NTSB
• Require EDRs (event data recorders, i.e. “black boxes”).
LEGAL ISSUES Automakers Google
● States currently allow only tesFng of self-‐driving vehicles.
● Federal government study completes only in 2017.
● California most likely state to allow public operaFon.
● NaEonwide support for assisFve technology (e.g, adapEve cruise control).
● Future technology can follow law changes hand-‐in-‐hand.
LIA
BILITY A
ND E
THIC
S Auto Insurance Industry
• Consumer Watchdog society to CA DMV on proposed regulations (03/11/2013) – “… gather only the data necessary to operate the vehicle and retain that data only as long
as necessary for the vehicle’s operation”.
• Poll conducted by Alliance of Automobile Manufacturers in 06/2013 highlighted the following consumer concerns: – 75% : operating software will be used to collect personal data – 70% : data will be shared with the government – 81% : hackers may gain control of vehicle
• FBI Strategic Issues Groups – report obtained 07/2014 – “get away” vehicles – Terrorist attacks – Criminal “multi-tasking”
Privacy & Security Concerns
~$200B : 2013 US Auto Insurance Premiums (Private & Commercial)
• Traditional insurance law – vehicles are insured not drivers.
• For Level 4 vehicles, does Private Motor Vehicle Insurance è Manufacturer Product Liability Insurance?
• Black Boxes è Easier fault determination?
13%
87%
2013 Premiums
Commercial
Private
5%
68%
25% 2%
2013 Costs & Profit
Profit
Claims
Overhead
Taxes
Source: The Insurance InformaEon InsEtute
Source: NaEonal AssociaEon of Insurance Comissioners
Can the auto insurance industry survive a 90/90 world?
90% fewer private cars è 90% drop in private premiums è $53B industry?
90% fewer accidents è 90% fewer costs in claims è 90% drop in premiums è$20B industry?
Source: The Eno TransportaEon FoundaEon
WIN
NE
RS / LO
SE
RS
** Data from mulEple sources: IHS and Market research
Losers Winners
Intelligent Traffic Highway & Mapping Ø IBM Ø Google
SDC Adopters Ø Google Ø Nissan Ø Audi Ø Mercedes Ø Toyota Ø Tesla
Integrated Component Suppliers
Ø BOSCH Ø ConEnental Ø Denso
Rental & Ride Sharing Companies
Ø AVIS, Hertz Ø UBER, LYFT
Auto Manufacturer Ø Reduce car sales Ø Low Margins
Taxi Services
Auto Insurance Companies Ø Reduce Premiums Ø New model for collision
coverage
Auto Service Industry Ø ConsolidaEon with few
survivors
Summary
BACKUP SLIDES
TH
E OP
PO
RTU
NITIE
S
TH
E EC
ON
OM
IC IM
PAC
T
Morgan Stanley esEmates that self-‐driving vehicles could deliver the following: Ø $1.3 Trillion in annual savings to the
U.S.Economy Ø $507 Billion in annual producEvity gains Ø $488 Billion in annual accident cost
reducEon savings Ø $158 Billion in annual fuel cost savings Ø $138 Billion in annual producEvity
savings from less congesEon
BE
NE
FITS O
F SE
LF-DR
IVIN
G C
AR
S
Ø More than 90 percent of all accidents are caused by human failure, according to the Network of Employers for Traffic Safety (NETS).
Ø The financial cost for US car accidents are esEmated to be $165 Billion every Year as par AAA 2008 report
Ø Average Car is Parked 96% of it lifespan Ø Massive Parking Area required for Car Parking Ø Self Driving cars would Save lives Ø Commute will become quicker and less stressful Ø Driverless Cars will be more fuel-‐efficient Ø Driverless car will make car-‐sharing program more popular Ø Driverless Car could bring billions of dollars in benefits
Self Driving Car Market Forecast
-‐2000000
0
2000000
4000000
6000000
8000000
10000000
12000000
14000000
2015 2020 2025 2030 2035 2040
SDC Volume forcast by IHS
SDC Volume forcast
$50,000.00
$10,000.00 $7,000.00
$1,000.00 $0.00
$10,000.00
$20,000.00
$30,000.00
$40,000.00
$50,000.00
$60,000.00
2010 2015 2020 2025 2030 2035
Cost of SDC Feature Add By IHS
Cost of SDC Feature Add
29%
20% 24%
27%
SDC Market Share Forcast by Region by IHS (2030)
USA
Europe
China
Others
RO
AD
MA
P
