The R&D Tax Credit Aspects of Driverless Cars
Driverless-Cars
The driverless car revolution is imminent.
Technology that was once considered science fiction is now coming down
the fast lane, carrying with it the ability to completely transform the
world of transportation. The present article will discuss the potential
benefits and challenges of self-driving vehicles and give an overview
of recent developments in the ongoing driverless car race. It will also
discuss how innovative companies can take advantage of R&D tax
credits as they open the way to a new era of mobility.
The
Research & Development Tax Credit
Enacted in 1981, the federal Research and
Development (R&D) Tax Credit allows a credit of up to 13% of
eligible spending for new and improved products and processes.
Qualified research must meet the following four criteria:
- New or improved products, processes, or software
- Technological in nature
- Elimination of uncertainty
- Process of experimentation
Eligible costs include
employee wages, cost of supplies, cost of testing, contract research
expenses, and costs associated with developing a patent. On
December 18, 2015, President Obama signed the bill making the R&D
Tax Credit permanent. Beginning in 2016, the R&D credit can be used
to offset Alternative Minimum tax and startup businesses can utilize
the credit against payroll taxes.
Rethinking
Transportation
Autonomous vehicles are bound to revolutionize
personal mobility, transforming our lives in several ways. They promise
to play a central role in the emergence of new city dynamics that can
better accommodate a growing urban population. In the words of UPS
Healthcare Strategy Director Wanis Kabbaj, “by a strange paradox, the more robotized
our traffic grid will be, the more organic and alive its movement will
feel.” Potential benefits include:
Safety: Though often
received with skepticism, research consistently shows that driverless
cars are safer than those driven by humans. Data from the Department of
Transportation (DOT) corroborate this idea, pointing out that 94
percent of car crashes are due to human choice or error.
Considered the “next revolution in
roadway safety”, autonomous vehicles are quite literally a
lifesaving technology. In 2015 alone, over 38 thousand people were
killed on U.S. roads and 4.4 million sustained traffic-related
injuries.
Traffic: Driverless cars
could significantly improve traffic flow and reduce fuel consumption.
Enhanced roadway safety means fewer accidents and less traffic
congestion. Additionally, the emergence of affordable self-driving
taxis promise to discourage car ownership, alleviating the overall
traffic. According to KPMG, autonomous vehicles could increase highway
capacity by up to 500 percent.
Fuel Efficiency:
Autonomous vehicles incorporate fuel-saving technology that optimizes
efficiency in acceleration, braking, and speed variation. Better
traffic conditions combined with enhanced fuel efficiency could help
reduce CO2 emissions produced by cars by as much as 300 million tons
per year, according to McKinsey.
Wellbeing and Mobility:
Taking the hands off the wheel can dramatically improve wellbeing. Long
commutes, which are often frustrating and exhausting, can become
restful and productive. Also, better traffic flow could considerably
reduce the time needed to get to a destination. McKinsey estimates that
commuters around the globe could save up to one billion hours every
day. Self-driving cars could also improve the mobility of
certain populations, such as the elderly, the unlicensed, and those
with disabilities.
Transportation Costs:
Studies anticipate a considerable reduction in car ownership, which
will be increasingly replaced by new mobility models, such as
“e-hailing” taxi alternatives. According to the Casualty Actuarial
Society, the cost per mile of a shared, driverless fleet car could be
80 percent smaller than the one of a personally owned vehicle when
driven 10,000 miles per year.
Economic Gains: A 2015
study by Morgan Stanley estimated that autonomous cars could contribute
$1.3 trillion in annual savings to the U.S. economy alone, with global
savings estimated at over $5.6 trillion. These gains represent a
combination of increased productivity due to shorter commutes, enhanced
fuel efficiency, along with savings from accident avoidance.
Urban Landscape: By
discouraging car ownership, driverless cars will also reduce the need
for parking space. This can have a major impact in what cities look
like, as parking spots currently account for 31 percent of urban land.
The Meaning
of Autonomous Driving
Business Insider (BI) Intelligence defines
self-driving vehicles as those containing features that allow them to “accelerate, brake, and steer a car's
course with limited or no driver interaction.” This definition
considers different levels of autonomy, from driving assistants all the
way to fully autonomous vehicles, which need no input from human
drivers.
In order to address the
diversity of self-driving technology, the National Highway Traffic
Safety Administration (NHTSA) recently updated its classification of
autonomous vehicles, which now echoes the five-level definition
proposed by the Society of Automotive Engineers (SAE) International.
Each level reflects “who does what,
when” as summarized by the following table:
Level 0
|
The
human
driver does everything;
|
Level 1
|
An
automated
system on the vehicle can sometimes assist the human driver
conduct some parts of the driving task;
|
Level 2
|
An
automated
system on the vehicle can actually conduct some parts of the
driving task, while the human continues to monitor the driving
environment and performs the rest of the driving task;
|
Level 3
|
An
automated
system can both actually conduct some parts of the driving
task and monitor the driving environment in some instances, but the
human driver must be ready to take back control when the automated
system requests;
|
Level 4
|
An
automated
system can conduct the driving task and monitor the driving
environment, and the human need not take back control, but the
automated system can operate only in certain environments and under
certain conditions;
|
Level 5
|
The
automated
system can perform all driving tasks, under all conditions
that a human driver could perform them.
|
Source: DOT, Federal Automated
Vehicles Policy – Accelerating the Next Revolution in Roadway Safety.
Sept. 2016.
The DOT further
distinguishes between levels 0-2 and 3-5. In the higher levels, the
automated system - not the human operator - is primarily responsible
for monitoring the driving environment. Therefore, cars that are
classified as levels 3-5 are denominated “highly automated vehicles” (HAV).
Taking into consideration
different levels of autonomy, Business Insider forecasts that the
self-driving market will experience a compound annual growth rate of
134 percent between 2015 and 2020, when there will be nearly 10 million
cars containing self-driving features on the road. Boston
Consulting Group estimates that autonomous cars might create a $42
billion market by 2025.
The
Technology Propelling Autonomous Vehicles
Self-driving vehicles are made possible by the
groundbreaking combination of four fundamental capabilities, namely,
mapping and localization, obstacle detection and avoidance, artificial
intelligence, and motion control. Though there can be different
approaches to the development of autonomous cars, key components
generally include:
LiDAR: Light detection and
ranging (LiDAR) is a surveying technology that measures distance by
illuminating a target with a laser light. It is used to create
high-resolution 3D maps of the vehicle’s surroundings. LiDAR modules
shoot out laser beams to determine the distance from and the profile of
nearby elements. LiDAR units are usually placed on top of autonomous
vehicles, where they can rotate and construct an unobstructed view.
Radar: Various radar units
are incorporated into autonomous vehicles in order to monitor the speed
of surrounding objects in real time. Working together with other
features of the car, such as inertial sensors, radars help avoid impact
by communicating when to swerve, apply the brakes, or prepare seatbelts
for impact. Some cars are also equipped with sonar technologies, as
complementary sources of data.
Cameras: High-powered
cameras contribute to mapping and localization. Smart cameras can
estimate distances to surrounding objects, read traffic signs, and
detect pedestrians.
Positioning Systems: The
ability to accurately locate itself is crucial to the safety and
functionality of an autonomous car. For this reason, HD maps with
centimeter precision and real-time capabilities incorporate a
localization layer that combines data from high-precision GPS receivers
and different sensor units.
Sophisticated Software:
Vehicle autonomy is based on its ability to process data and adjust to
traffic conditions. Complex algorithms are used to analyze incoming
data in real time as well as to model the behavior of surrounding
agents, such as other drivers, pedestrians, etc. In addition,
machine-learning capabilities allow self-driving systems to optimize
their responses based on previous experiences and on “observ
ing” and learning from other cars.
The
15-point Safety Checklist
In September 2016, the DOT unveiled the Federal
Automated Vehicles Policy, laying a path for the safe testing and
deployment of driverless cars. The document includes a 15-point
checklist that outlines safety expectations for the emerging
technology. It is considered to be an important step in balancing the
necessary safety oversight and the commercial interests at hand.
Overall, the guidelines, which were entitled “Accelerating the Next Revolution in
Roadway Safety”, were seen as an indication that federal
regulators endorse the potential benefits of driverless cars and that
innovative companies should anticipate a scenario of collaboration
rather than overregulation.
As demonstrated below, the 15
assessment areas listed by the DOT are some of the most important areas
for innovation in driverless car technology. Companies engaged in
finding new and improved ways to overcome existing challenges in each
one of these areas should take advantage of R&D tax credits to
increase their chances of success.
1. Data Recording and Sharing
The document encourages carmakers
not only to collect and store data for testing and operational purposes
but also to develop strategies for sharing information that could “help to accelerate knowledge and
understanding of HAV performance, and could be used to enhance the
safety of HAV systems and to establish consumer confidence in HAV
technologies.” In cases of crash reconstruction, data concerning
the event and the performance of the vehicle should be readily
available for retrieval by regulators.
2. Privacy
In an effort to protect
consumers’ right to privacy, the DOT points out that car owners should
not only be made aware of the data being collected, but also be given
the opportunity to decline authorization for the collection of personal
information, such as geolocation, biometric, and behavioral data. The
document also underlines the importance of using data only in “ways that are consistent with the purposes
for which it was originally collected.” Accountability,
integrity and access, data security, and de-identification are also
points of concern.
3. System Safety
The ability to safely
respond to system malfunctions should be a priority in vehicle design.
Safety considerations should include potential electrical, electronic,
or mechanical malfunctions as well as software errors. Design decisions
should be tested, validated, and verified to demonstrate the vehicle’s
overall safety in problematic situations.
4. Vehicle Cyber Security
We are currently witnessing an
unprecedented multiplication of cyber crimes. The DOT urges
carmakers to reduce cyber security vulnerabilities by using
identification, protection, detection, response, and recovery
functions. Risk management decisions, quick response to threats, as
well as the ability to learn from cyber security events should be a
priority in vehicle system design. The Agency further encourages
companies to share information on cyber security with their
counterparts, as a means to accelerate industry-wide advances.
5. Human Machine
Interface
Emerging
technology brings new complexities to the interactions between vehicles
and humans. The DOT underlines that automakers should consider how
vehicles will communicate status and intentions not only to the
occupant but also to the environment around it, including pedestrians,
cyclists, and other cars. This is particularly important in
semiautonomous systems that switch between autopilot and human control.
Human machine interactions should also accommodate people with
disabilities.
6. Crashworthiness
Driverless cars must
meet NHTSA’s crashworthiness standards. Manufacturers are also
encouraged to develop new occupant protection systems “that use information from the advanced
sensing technologies needed for HAV operation to provide enhanced
protection to occupants of all ages and sizes.” In the case of
non-occupied automated cars, automakers should make sure they provide
geometric and energy absorption crash compatibility with existing
vehicles.
7. Consumer Education
and Training
The DOT encourages carmakers to
deploy wide-ranging education and training programs targeted at
employees, dealers, distributors, and consumers. The objective of such
initiatives is to make sure that technologies are used properly,
efficiently, and in the safest manner possible. Besides traditional
on-road and on-track experiences, the Agency encourages the use of
innovative approaches to education, including virtual reality.
8. Registration and
Certification
In addition to requiring that
software updates and new driverless features be submitted to the NHTSA,
the document underlines the importance of providing up to date,
on-vehicle information of key HAV capabilities to human drivers,
owners, and occupants.
9. Post-Crash Behavior
After a crash,
manufacturers must demonstrate their vehicle’s ability to be safely
reinstated into service. Damages to critical control systems should
prevent further operations in autonomous mode.
10. Federal, State and
Local Laws
Driverless cars must comply with
all applicable federal, state, and local laws. However, they must also
be able to identify “plausible cases” in which violations of the law
are necessary to avoid accidents (such as crossing double lines to
avoid a broken-down vehicle). Carmakers are encouraged to record data
that prove that actions taken by the vehicle’s system were
safety-promoting. They should also envision the possibility of changing
legal landscapes, which must be incorporated into software updates.
11. Ethical Considerations
Baring in mind that driving
decisions can have ethical dimensions or implications, the DOT
encourages manufacturers to work together with regulators, drivers, and
passengers to develop problem-solving algorithms that make sure that
ethical judgments are made “consciously
and
intentionally.” This is particularly pressing in instances
where there is a conflict between safety, mobility, and legality - the
three major principles of vehicle operation.
12. Operational Design Domain
Carmakers should disclose
the Operation Design Domain (ODD) for each system capability, including
information such as 1) roadway types on which the HAV system is
intended to operate safely; 2) geographic area; 3) speed range; 4)
environmental conditions in which the HAV will operate; and 5) other
domain constraints. Automakers are also encouraged to have documented
processes and procedures for the assessment, testing, and validation of
the system’s capabilities.
13. Object and Event Detection
and Response
Autonomous vehicles must be
able to detect and respond to any circumstance that is relevant to the
immediate driving task. Manufacturers should demonstrate that their
vehicles are capable of safely interacting with other cars,
pedestrians, animals, and cyclists as well as dealing with unexpected
conditions, such as temporary work zones, manually directed traffic,
etc.
14. Fall Back (Minimal Risk
Condition)
When faced with
technological malfunctions, HAVs should be able to safely transition to
a minimal risk condition. Systems that allow for human driving should
consider the potential driver’s condition before switching to human
control, taking into account the possible influence of alcohol,
drowsiness, etc.
15. Validation Methods
The DOT acknowledges the
wide variety of technologies used in driverless cars and urges
manufacturers to develop the necessary tests and validation methods to
ensure high levels of safety. Simulation, test track, and on-road
tracking should integrate a multidimensional approach to testing HAV
systems.
The safety guidelines above are
expected to work as a foundation and a framework upon which future
governmental action will occur. In addition to listing safety
assessment areas, the document also encourages states to create uniform
rules to address autonomous vehicles, as the existing patchwork of
state laws is seen as a potential hurdle to the safe yet timely
deployment of driverless cars. In a section entitled “Model State
Policy”, the DOT states: “a
manufacturer should be able to focus on developing a single HAV fleet
rather than 50 different versions to meet individual state requirements.”
Hitting the
Road: Recent Developments in the Autonomous Car Race
Despite major technological breakthroughs,
self-driving under real-world conditions still poses significant
challenges. A growing number of automakers and technology companies are
working to overcome these hurdles, in an exciting race for the
commercialization of autonomous vehicles. The following sections
explore the work of these groundbreaking companies, which exemplify the
kind of innovative efforts that could benefit from R&D tax credits.
Tesla
In October 2016, Tesla
announced that all its new cars would have hardware for “full self-driving capabilities.” The full autonomy
update includes 8 cameras with 360-degree viewing at up to 820 feet of
distance, 12 ultrasonic sensors that can detect both hard and soft
objects, as well as a forward-facing radar unit to increase visibility
through rain, fog, and dust.
According to CEO Elon Musk,
the new hardware is 40 times more powerful than the previous Tesla
computer. The automaker is equipping its vehicles with Nvidia’s Drive PX
2,
a supercomputer that translates sensor data into driving commands
and uses deep learning to teach the car to handle itself. For now,
Tesla has purposely avoided using LiDAR sensor technology. While some
experts corroborate this choice, pointing to the costs and
vulnerabilities of LiDAR, others believe these sensors are essential to
guaranteeing 100 percent safety. Potential future improvements to
Tesla’s system include CMOS image sensors from longstanding Tesla
partner Panasonic. The Japanese company has expressed interest in a
collaboration for the development of very high-speed sensing
technology.
Though allegedly equipped
for complete automation, full self-driving capabilities are not
available in Tesla’s cars just yet. As stated by the automaker’s
website “self-driving functionality
is dependent upon extensive software validation and regulatory
approval, which may vary widely by jurisdiction.” However,
recent software updates have been seen as important steps towards the
long-term goal of producing fully autonomous, self-driving cars as well
as the more near-term objective of fully activating the Enhanced
Autopilot, which is expected to include functions such as speed
matching with surrounding traffic, lane keeping, automatic lane
switching, freeway merging and exit, along with self-parking and
summoning on startup.
In December 2016, Tesla
reassured its customers that progress is being made in the refinement
of a new vision neural net system, which is the basis of various
autonomous features, including object identification and avoidance.
Musk has pledged to release the Enhanced Autopilot and perform a
cross-country test of a fully-automated, driverless vehicle before the
end of the year. Even though this ambitious plan often inspires
skepticism, having access to massive amounts of data collected by Tesla
cars around the globe (over 222 million miles in Autopilot mode) may be
a game-changing advantage.
Uber
In September 2016,
ride-hailing service Uber crossed a major milestone by introducing
autonomous cars in Pittsburgh, Pennsylvania. The self-driving vehicles,
which are available to select users in limited areas of the city, rely on
cutting-edge technology, including 20 cameras that watch for slowing
vehicles, pedestrians, and other obstacles; LiDAR modules that shoot
1.4 million laser points per second to create a 3D map of the
surroundings and detect obstacles in blind spots; as well as a cooling
system to prevent overheating. A safety driver and an engineer
occupy each car to take control in the event of failures. Uber has
given several indications of its determination to build a fleet of
self-driving vehicles. However, contrary to most players in the
driverless-car race, the startup has decided to outfit cars with
autonomous driving kits rather than developing new vehicles from
scratch. Uber recently established a $300 million partnership with
Swedish-based carmaker, Volvo, to develop a fully autonomous car ready
for the road by 2021, as reported by Bloomberg. To this end, the
company has also hired various robotics experts from Carnegie Mellon
University.
In December 2016, Uber
launched its second pilot program in San Francisco but shut it down due
to licensing problems with California’s DMV. The cars were sent to
Arizona, where the company plans to resume the program. Additionally,
Uber is involved in the development of self-driving freight trucks. In
August 2016, it acquired San Francisco-based Otto for an estimated $680
million. Just two months after that, a self-driving Otto truck made a
120-mile run to deliver beer.
Ford
In August 2016, Ford
announced its intention of having high-volume, fully autonomous
vehicles in commercial operation by 2021. The automaker aims to design
SAE level 4 vehicles for use in urban car-sharing and ride-hailing
fleets, which will operate without a steering wheel or pedals. To
achieve this ambitious objective, Ford has invested in startups,
increased its Silicon Valley team, and more than doubled its Palo Alto
campus. The company has also pledged to triple its autonomous vehicle
test fleet in 2017, bringing the total number to 90 cars. The testing
vehicles will help develop a “more targeted field of vision” and
improve on Ford’s virtual driver system, which includes cameras, radar
and LiDAR sensors; algorithms for path planning; computer vision and
machine learning; 3D maps; and advanced electronic systems.
Recent tests have included the operation of autonomous vehicles in the
snow and at night, in complete darkness. Key investments and
collaborations include Velodyne, Silicon Valley-based developer and
manufacturer of LiDAR technology; SAIPS, Israel-based computer vision
and machine learning company; Nirenberg Neuroscience, a machine vision
company that has created a powerful machine vision platform for
performing navigation, object recognition, facial recognition and other
functions; and Civil Maps, Berkeley, California-based developer of
high-resolution 3D mapping capabilities. Ford has recently
unveiled a new model of its self-driving car Fusion, which featured a
more “traditional” automobile look, made possible by the scaling down
of hardware, particularly the bundle of sensors previously located on
the roof of the vehicle. The new generation of autonomous Ford cars has
a sensing range that extends the equivalent of two football fields in
every direction. Ford has recently registered a patent for a
self-driving car equipped with a drone. The idea is to use the drone to
map the surrounding areas beyond the reach of the vehicle’s sensors.
General
Motors
American automaker GM has
taken significant steps towards developing autonomous vehicle
technology. In January 2016, the company invested $500 million in the
ride-hailing startup Lyft, in what was reported as the biggest single
Detroit-Silicon Valley deal so far. The newly formed partnership aims
to create an integrated network of on-demand autonomous vehicles. It
illustrates the increasing interest of automakers in ride-sharing
programs as a doorway to autonomous vehicle deployment. GM has also
recently started a car-sharing company, Maven, which offers on-demand
cars to drivers via an app. In March 2016, GM acquired Silicon
Valley-based Cruise Automation for about $600 million. The goal is to
combine the startup’s innovative self-driving technology with GM’s
manufacturing capabilities, allowing for the deployment of autonomous
vehicles at a mass scale. In June, the automaker began testing
autonomous Chevrolet Bolt EVs in San Francisco and Scottsdale, Arizona.
With 40 autonomous vehicles currently being tested, GM plans to expand
its testing fleet to public roads in metro Detroit, which will be key
to advancing autonomous technologies in harsh winters. By
producing the next generation of all-electric, self-driving cars in
Orion Township, Michigan, GM intends to become the “first high-volume auto manufacturer to
build fully autonomous vehicles in a mass production assembly plant.”
Bosch
German multinational auto
supplier Bosch has worked on the development of new components to
support full autonomy. With a global team of nearly 2,500 engineers
dedicated to advancing driver assistance systems and automated driving
technology, Bosch’s clients include Google, Tesla, and Porsche. The
company focuses primarily on sensor technology – particularly radar,
video, and ultrasonic sensors, vehicle architecture, as well as
actuators and their integration into the vehicle. Bosch has
recently developed fully autonomous prototypes by incorporating its
technology into Tesla Model S sedans – the company underlined, however,
that its system-based approach could be used in any vehicle. The
prototype has level 4 autonomy, made possible by six radars, six LiDAR
modules, one stereo video camera, and one high precision GPS. The
company has also expressed interest in combining autonomy and
connectivity, linking vehicles to the Bosch IoT cloud and allowing for
major increases in productivity both in household functions and
logistics operations.
Google
In December 2016, Google
transformed its driverless car unit into a new, independent company
that will operate under parent company Alphabet - a move that could indicate the growing commercial
viability of driverless technology. The newly created Waymo has
signaled that it intends to incorporate autonomous-driving technology
into typical cars, instead of developing new vehicles.
Focusing on the scalability
and optimization of self-driving technology, the company is engineering
and manufacturing all of its hardware in-house. Waymo CEO John Krafcik
recently announced that the company has brought down the cost of LiDAR
sensors by 90 percent – a change that could significantly reduce the
overall price of autonomous cars, which often have LiDAR as
their most expensive component. In addition to cutting costs, Waymo
claims to have used more than 2 million test miles to create sensors
that are more resistant to vibration and extreme temperatures.
The company is expected to
soon begin testing a fleet of Fiat Chrysler Pacifica minivans equipped
with its self-driving technology, which includes a suite of LiDAR
sensors that operate at short, medium, and long ranges as well as eight
enhanced camera modules specially designed for challenging lighting
conditions. Waymo intends to market driverless technology for
various applications, including personal transportation, public transit
systems, ride hailing, and trucking.
Mercedes-Benz
In October 2015, a
Mercedes-Benz big-rig became the first heavy-duty truck to drive
semi-autonomously on an open highway. Equipped with Daimler AG’s
Highway Pilot, which consists of a system of radars, sensors, and
cameras, the truck featured various safety functionalities, including
brake assist, active cruise control, and drowsiness detection.
Mercedes-Benz’s goal is to transform regular trucks into self-driving,
intelligent ones by incorporating its innovative system, which should
be available for commercial applications by 2020.
Mercedes recently announced
the upcoming release of its new Drive Pilot, a semi-autonomous system
that controls steering and speed while remaining under the driver’s
supervision. The new technology will represent a major improvement over
the company’s existing Drive Pilot: while the current version can
handle only 20 percent of driving tasks, the new system will be able to
take charge of up to 80 percent of such tasks. In addition to greater
capabilities, Mercedes has enhanced human-machine interaction through
an innovative feature that allows drivers to decide and communicate how
much input they want to give. The new Drive Pilot adapts to how much
steering force is used: while a light touch signals that the system
should do most of the work, a firmer hand gives the control back to the
driver. Mercedes underlines, however, that Drive Pilot consists of an
assist system, not a fully autonomous one. For this reason, the company
incorporated features designed to avoid overreliance – for instance, it
requires a driver’s response every few seconds, depending on the road
conditions and speed. Failing to acknowledge these requests activates a
“controlled-but-determined emergency stop.” Mercedes recently
announced a partnership with American technology company Nvidia. Having
worked together on deep learning and artificial intelligence for the
last three years, the two companies intend to roll out a self-driving,
artificially intelligent car by 2018.
BMW
Working in partnership with
Intel and Israeli computer vision company Mobileye, BMW intends to have
40 self-driving test vehicles on the roads by the second semester of
2017. As part of the Project iNext, the German carmaker aims to release
an all-electric car with autonomous functions by 2021 and a fully
autonomous vehicle by 2025. BMW recently unveiled the BMWi Inside
Future, an innovative concept for the interior of its autonomous cars,
which is built around the idea of connectivity and includes a new
user-interface named HoloActive Touch system – a “free-floating”
display that works via gestures, not touch. Other features are a
folding steering wheel that retreats when in autonomous mode, reclining
seats, and a screen for entertainment in the backseat. The futuristic
design points to a hallmark of driverless innovation – in the words of
Klaus Frohlich, a member of BMW's board of management, "it’s not about just being driven but about
the experience."
The list of companies
currently racing to build self-driving cars also includes Japanese
automakers, such as Toyota, which has invested $1 billion in a research
institute focused on artificial intelligence and robotics technology;
Nissan, which is gradually incorporating autonomous capabilities into
its cars through the innovative ProPILOT system; and Honda, which
recently unveiled a prototype for a self-balancing, autonomous
motorcycle. Audi also has plans to roll out level 4 autonomous
cars by 2020. The German auto giant has been testing its “piloted
driving” technology, which doesn’t rely on pre-established
instructions, but rather observes and learns from other cars (with
human drivers), adapting to different contexts and is thus riding more
“naturally.” Hyundai and French multinational PSA Group are also
engaged in driverless technology research. Apple has repeatedly
expressed its interest in using machine learning and automation in
transportation. There is, however, little information as to the nature
of the company’s efforts - building a physical product or developing an
autonomous system to be used by other cars.
Michigan:
Paving the Way for Driverless Innovation
In December 2016, Governor Rick Snyder signed a bill
that allows autonomous cars on Michigan’s public roads, no back driver
required. The legislation has been referred to as the “most permissive self-driving car laws in
the country” as it allows testing of vehicles without steering
wheels or pedals (which are prohibited in other states, such as
California). The idea is to put Michigan at the forefront
of driverless vehicle development, reaffirming its position as the
epicenter of automotive innovation. According to the director of
Michigan’s Department of Transportation Kirk Steudle, over 75 percent
of companies engaged in autonomous vehicles R&D already have
locations in Southeast Michigan, and the state wants to “encourage them to stay put.” The
progressive legislation further wants to welcome new businesses – in
Steudle’s words, “we welcome any
company to come on in and operate on the roads. Show us what you got.”
Tech
companies and automakers have welcomed the new legislation, which
is already attracting new businesses. Uber has recently announced that
it will open an autonomous-vehicle research center in Wixom, Michigan,
and General Motors expressed its intention to develop and build
autonomous Chevrolet Bolt EVs at the Orion Township assembly plant.
Other initiatives to advance
driverless innovation include University of Michigan’s Mcity, a 32-acre
simulated city for autonomous car testing. Located in the Ann Arbor
campus, Mcity replicates the urban environment, with traffic jams,
pedestrians, angled intersections, obstructed views, and even a
four-lane highway. The testing center has attracted various
startups that take advantage of the initiative’s TechLab, a learning
incubator where U-M students contribute to early-stage technology
development. Examples of West Coast companies that have recently joined
MCity include driving analytics startup Zendrive, which uses smartphone
sensors to measure drivers’ behavior; PolySinc, developer of an
operating system for fully autonomous vehicles; and Civil Maps, creator
of a 3-D mapping technology that transforms sensor data into map
information specially targeted at driverless cars.
The state will soon have
another major testing site, the proposed American Center for Mobility.
Located in a World War II bomber factory and former GM plant in the
Ypsilanti Township, the $80 million, 335-acre project will feature a
2.5-mile highway loop, which should be concluded by December 2017, as
part of the first phase of constructions.
Michigan is home to various
innovative companies that are working on driverless technology.
Detroit-based automotive supplier Roush has developed proprietary noise
and vibration technology as well as analytical models that predict
product performance and determine cost-effective solutions in varying
environmental conditions. In 2015, Roush was reported as one of
Google’s partners in its self-driving car efforts. The state is
also home to FEV North America, a leader in the development and testing
of electronics systems and subsystems for Advanced Driver Assisted
Systems (ADAS). FEV has developed cyber security solutions capable of
protecting self-driving cars from hackers. Its Cyber Security Gateway
works as a firewall between interface threat entry points (such as
WiFi, Bluetooth, etc.) and the vehicle. It can be used as a standalone
or integrated solution connected to the car’s communication bus.
Michigan is also home to Techstars Mobility,
the first startup accelerator program in North America to focus on the
future of mobility, automotive, and transportation. Working from
downtown Detroit, the accelerator has invested in over 22 mobility
companies since 2014. Techstars’ 2016 class of startups included New
York City-based Braiq, which uses biosensors to bridge the gap between
human preferences and artificial intelligence as a means to personalize
the driving style of autonomous vehicles. Braiq’s innovative and
ambitious goal is to teach self-driving cars how to respond to human
emotion by monitoring passengers’ biosignals, such as brain activity,
eye movement, facial expressions, and heart rate.
Conclusion
Autonomous vehicles promise to revolutionize
transportation. Potential benefits include increased road safety,
better traffic flow, higher fuel efficiency, shorter commutes, among
many others. Various companies are engaged in innovative efforts to
overcome the technological challenges to the safe and widespread
deployment of driverless cars. R&D tax credits can help pave the
way to the commercialization of this revolutionary technology.
