
[Music]
have you ever been on your way to work
or school only to see a traffic crash at
the side of the road life slows down for
a moment to see what happened see if
anybody’s hurt and to see the activities
of first responders what we often forget
about everyday that although small
there’s a slight risk of being injured
or killed in an auto crash the World
Health Organization lists road traffic
injuries as the eighth leading cause of
death worldwide
currently over 1.25 million fatality
occur on the world’s road and over 50
million individuals are non fatally
injured by many measures this is a
massively undertreated global health
crisis what have we been doing to think
about this driver error has long been
regarded as a critical factor in over
90% of these accidents so more recently
we’ve turned and begin to talk about
automation to fix this problem we first
focused on automation to make driving
easier like automatic transmissions
power steering and cruise control more
recently we’ve looked at automation for
crash protection and prevention
technologies such as electronic
stability control airbags for collision
warnings and more recently automatic
emergency braking we hear in the news
all the time the focus of on automation
to make driving easier perhaps relieving
the burden altogether and not so far in
the distance some cars will indeed
become driverless the context here is
that robots can be better drivers than
us but the question becomes how safe
will this all be just look at the roads
of today who would have guessed that
automating simple things such as
shifting gears would have made it so
much easier to pick up smartphone with
increased automation comes increased
responsibility and driverless does not
mean human las’ integrating the
expertise of humans and machines has
long been understood to be critical for
safety
driverless vehicles will need human
support perhaps Catan perhaps it afar
for sometime in a pursuit of automation
we very much been letting the technology
lead the way and forgetting about to
some degree the behavior of the driver
behind the wheel
well we’re once on a pathway to safer
roads in more recent years our roads
have turned more risky many you’d like
to blame this undistracted drunk drowsy
and drugged driving however we may need
to look beyond the driver themselves
perhaps a critical elements of vehicle
technology the infrastructure education
and other factors that may be coming
together to impact safety so as we look
to the increasing amount of automation
on roads how is this all going to work
as a society we’ve been long enamored
with automotive technology we like to
touch it feel it see it automated and
our collaborative hands-free driving is
here today technologies such a Tesla
autopilot Cadillac super cruise Volvo’s
pilot assist here’s a sense of how it
looks which was once Jetsons like
science fiction on our roads as a
current our roles of drivers changing
realistically our automated future is
not going to be here tomorrow
it takes about 18 years to change over
about 85% of the fleet here in the
United States so for a long time to come
if not forever there will be a mix of
driverless automation assisted in plain
old human drivers on the road what we
almost always forget is that our
infrastructure was really designed and
built for human drivers realistically
this is not going to transform
seamlessly into a system to enable the
future of mobility when we can’t even
afford to paint the lines pave the
potholes or fix the bridges in an
infrastructure where is still expected
to travel with Model Ts
I spent years studying human behavior
and it worries me how many individuals
and organizations don’t understand the
level of complexity
in the transformation of our system that
is required to enable safe convenient
and enhanced mobility technologists
automakers politicians and drivers see
the world through different lenses who’s
right who’s wrong or is a blend of all
views needed
so one disconnect we see today is that
many of the safety technologies that
we’re buying at the dealership are not
being enabled the concepts make sense
but something’s not right looking here
at some data from the Insurance
Institute for Highway Safety we can look
at lane departure warning technologies
that were designed and built to mitigate
run off the road crashes one of the
world’s most deadly in cars equipped
with these systems they’re only enabled
45% of the time when they show up at a
dealership for service our work at MIT
suggests that false alarms confusion
over where and when and why these
systems are there to support you as part
of the cause we believe that education
is critical to understanding when and
where to use these technologies it
appears that many don’t realize that
with automation comes a responsibility
for increased education and systems to
promote proper use of the technology
I for one often get into my car with
backup cameras and remote sensing
backing up out of a parking spot whether
that’s parallel or perpendicular and
never fully inspecting the surroundings
around me I think many of us experienced
that today and we forget that this was
automation all created to assist us as
opposed to replace human sensing so
we’ve been hearing tales for some time
now of how automated vehicles are going
to enable safe more efficient
transportation however several recent
events including the tragic uber
fatality and several autopilot related
crashes illustrate that risks remain
factors related to these crashes suggest
that perhaps we need to be thinking more
about some of the softer elements of the
system in essence looking beyond the
core automation technology itself
technologists would say what’s missing
automakers tell us there are
the road to doing it better politicians
can we trust industry to do this for us
on the other hand us as drivers when’s
this driverless system going to be
available to take me to work
take my ailing parents to the doctor or
drop my kids off at school by itself
making my daily life a lot easier so
it’ll be more than fine tuning the
robots it’s gonna be more than building
a better car it’s going to be more than
moving the driver to the passenger’s
seat
our teen regulatory efforts will need to
be cast aside and as a society we’re
gonna need for an evolution versus a
revolution of the robots on our roads as
they will more likely be able to help
get you to safely to work
before they’re gonna be able to take
your parents that a doctor or your kids
to school on their own when we think
about these systems there are some tough
decisions to make because everything
around how we live and move may need to
be upended to get this right how safe is
safe enough just safer than the average
Joe or 50% safer more questions of who
to avoid and win a child playing in the
road or an older couple walking down the
sidewalk should we be thinking about
building automotive vehicles that aim to
replace human drivers it’s going to be
difficult to prevailed an artificial
system with the seemingly irrational
behavior of human drivers we often
ignore rules of the road speeding
squeezing through yellow lights or
making instantaneous movements based
upon almost unexplainable movements of
other cars around us when we think about
this how do we begin to develop systems
that we trust efficiently that we won’t
walk in front of it because we know
they’ll stop or we won’t merge in front
of because we know they’ll yield
expectations of the system are critical
to building trust and Trust is built on
experiences education may be a very
critical element in developing an
understanding of where the advantage of
automation are and where and when we may
neither take the wheel driver monitoring
you highlight in the news the thing that
we need to think about connecting to the
automation so it is aware of the drivers
attention to the world around us and
capable of forming the driver when
attention wanes
infrastructure improvements are
critically needed so that the automation
isn’t that spoofed by seeming failures
in our road infrastructure these and
other changes will not be easy and will
not occur without some stimulus we don’t
have a unified strategy for automating
driverless vehicles here in the United
States or much the rest of the
industrialized world federal policy
discussions are verily centered on what
I would consider is minor changes
today’s decades-old regulatory framework
that was created before microprocessors
and software ruled the world
without federal leadership we have 50
states heading in 50 different
directions clearly we need to develop
the foundations to enhance the system
and accelerate the process of saving
lives
politicians can make radical change
difficult in the timeframes needed to
keep up with this technology so I’d
argue that an alternative approach is
really needed automated vehicles may
seem radical and new but automation is
being leveraged by industries worldwide
in countless other settings there are
case studies from these industries that
we can leverage to build a foundation
for this safe mobility system that we
all desire for over 50 years the Federal
Aviation Administration has been working
to transform aviation system into the
trusted system we know today the
aviation industry has long known that is
the relationship between automation and
humans that is critical to safe flight
that’s why autopilot in particular is
licensed under the supervision of a
human operator the FAA has stressed
pilot education and training an
incredible investment in networked
infrastructure coordinated control and
tracking of aircraft data sharing and a
safety culture
oh this I see the safety culture aspect
perhaps being the most important the
automotive industry has long competed on
safety and perhaps a collaborative
safety culture is the critical element
to get organizations to collaboratively
work together to solve some of the most
difficult problems in education
infrastructure
and on a coordination of the automated
vehicles on our roadway the National
Transportation Safety Board is charged
with investigating major safety related
events related to the transportation
infrastructure like planes trains and
automobiles crashes so by default this
agency already has a role in
investigating the recent uber and Tesla
incidents but changes this an agency are
needed as well because cars by their
sheer number and the complexity of the
operating environment are going to crash
far more than planes and change trains
society can’t afford to wait a year for
an investigative report with from an
entity with the little enforcement power
when hundreds of thousands of automated
vehicles are on our roads and the
software systems fueling those are
changing by the day so a well-funded
nonpartisan safety focused scientific
entity really needs to be focused on
advancing the insights needed to
understand how to fuel the policies and
safety developments of the future if the
International Transportation Safety
Board remains as it is one day it will
not have the resources to respond to a
787 crash or the ability to impart
insights learnt fast enough to save
lives would you want one of your family
members or loved ones involved in a
crash as a result of a software
limitation that was known the Food and
Drug Administration has been regulating
the balance of benefits and costs and
medical device and pharmaceutical
discovery for decades with so many lives
lost to traffic collisions we need to
embrace technology such as automation
they can move the needle in a positive
safety direction while managing the
risks much likely manage the side
effects of a life-saving drug one has
the question should we be allowing way
moe to deploy tens of thousands of
automated vehicles on our roads without
some public understanding of the
cost-benefit risk model management plan
and tracking some system for side
effects that may occur after deployment
the clinical trial structure provides us
with a clear framework for testing
innovations in unknown
areas and stages we could test many
vehicle automation technologies in
similar way test the system components
check for integrated safety check for
efficacy on the road and collect more
data to garner the findings in short
requiring that companies demonstrate
they can walk before anybody is allowed
to run a strong post market to program
to ensure we have gotten this right or
making the changes needed maybe the most
important piece of the puzzle so I’m not
standing here and saying that we can’t
move fast I think we can a clinical
trial structure in the automotive
industry to the last days or weeks as
opposed to the years it takes in the
clinic is data coming from vehicles
comes much faster the key here is we
need to keep think carefully and
stepwise as a society before we allow
the technologists too much Liberty to
treat us as non-consenting test subjects
as such something new really may be
needed here if the government is going
to set stage by itself we need to think
big I can easily envision something as
transformative as a creation of a
department is in the context of Homeland
Security a new mission a new budget for
a new technology realistically this is
not easy our government does not move
fast nor too many of the rest of the
world nor is the technical skill needed
to understand this car technology really
there at the governmental level as
something we really need to think new
radical different perhaps in the context
of a private utility with technologists
and automakers leading the way but the
public sector charged with protecting
benefits and societal goods like safety
the context of working together in a
public-private partnership where
automakers and policy makers come
together to draft the policies that are
needed educational approaches testing
programs infrastructure changes ethical
dilemmas and answering questions such as
how safe is safe enough all under the
air of transparency affected has been
missing to date in essence trying to
create the alignment between stake hold
so that we can fuel the research and
efforts that are needed on the softer
side of this technology this is all
critical to building our trust in the
future of automated systems the status
quo is just not going to work when
robots are causing traffic jams
automated driverless vehicles are our
future but we can’t manage this
technology with dated measures that were
built before the era of software we know
that the free market left on its own is
not going to manage than weigh the cost
and benefits well by itself the biggest
problem here may be the will of all the
stakeholders to come together along the
pathway I described or an alternate
route to make the changes needed the
future of personal mobility and safety
is exceedingly bright when coupled with
the technology at hand and bumps on the
road that are going to occur like recent
incidents can be managed we’ve been here
before it was collaboration that got us
the moon and back safely politicians
wanted it companies wanted it
constituents wanted it so they made it
happen perhaps that’s the best framework
to be thinking about the modernization
and automation of our mobility ecosystem
in the long run automated vehicles will
enhance safety the question is can we
collaboratively work together to take
the reins of this public health crisis
and collaboratively strategically move
the safety faith fight forward I really
hope we can as too many people are dying
on the world’s roads every day and I’d
like my kids riding and driving in safer
roads it’s important we get this right
and do it quickly if we can automate it
in driverless vehicles will be the most
critical life-saving technology of the century
have you ever been on your way to work
or school only to see a traffic crash at
the side of the road life slows down for
a moment to see what happened see if
anybody’s hurt and to see the activities
of first responders what we often forget
about everyday that although small
there’s a slight risk of being injured
or killed in an auto crash the World
Health Organization lists road traffic
injuries as the eighth leading cause of
death worldwide
currently over 1.25 million fatality
occur on the world’s road and over 50
million individuals are non fatally
injured by many measures this is a
massively undertreated global health
crisis what have we been doing to think
about this driver error has long been
regarded as a critical factor in over
90% of these accidents so more recently
we’ve turned and begin to talk about
automation to fix this problem we first
focused on automation to make driving
easier like automatic transmissions
power steering and cruise control more
recently we’ve looked at automation for
crash protection and prevention
technologies such as electronic
stability control airbags for collision
warnings and more recently automatic
emergency braking we hear in the news
all the time the focus of on automation
to make driving easier perhaps relieving
the burden altogether and not so far in
the distance some cars will indeed
become driverless the context here is
that robots can be better drivers than
us but the question becomes how safe
will this all be just look at the roads
of today who would have guessed that
automating simple things such as
shifting gears would have made it so
much easier to pick up smartphone with
increased automation comes increased
responsibility and driverless does not
mean human las’ integrating the
expertise of humans and machines has
long been understood to be critical for
safety
driverless vehicles will need human
support perhaps Catan perhaps it afar
for sometime in a pursuit of automation
we very much been letting the technology
lead the way and forgetting about to
some degree the behavior of the driver
behind the wheel
well we’re once on a pathway to safer
roads in more recent years our roads
have turned more risky many you’d like
to blame this undistracted drunk drowsy
and drugged driving however we may need
to look beyond the driver themselves
perhaps a critical elements of vehicle
technology the infrastructure education
and other factors that may be coming
together to impact safety so as we look
to the increasing amount of automation
on roads how is this all going to work
as a society we’ve been long enamored
with automotive technology we like to
touch it feel it see it automated and
our collaborative hands-free driving is
here today technologies such a Tesla
autopilot Cadillac super cruise Volvo’s
pilot assist here’s a sense of how it
looks which was once Jetsons like
science fiction on our roads as a
current our roles of drivers changing
realistically our automated future is
not going to be here tomorrow
it takes about 18 years to change over
about 85% of the fleet here in the
United States so for a long time to come
if not forever there will be a mix of
driverless automation assisted in plain
old human drivers on the road what we
almost always forget is that our
infrastructure was really designed and
built for human drivers realistically
this is not going to transform
seamlessly into a system to enable the
future of mobility when we can’t even
afford to paint the lines pave the
potholes or fix the bridges in an
infrastructure where is still expected
to travel with Model Ts
I spent years studying human behavior
and it worries me how many individuals
and organizations don’t understand the
level of complexity
in the transformation of our system that
is required to enable safe convenient
and enhanced mobility technologists
automakers politicians and drivers see
the world through different lenses who’s
right who’s wrong or is a blend of all
views needed
so one disconnect we see today is that
many of the safety technologies that
we’re buying at the dealership are not
being enabled the concepts make sense
but something’s not right looking here
at some data from the Insurance
Institute for Highway Safety we can look
at lane departure warning technologies
that were designed and built to mitigate
run off the road crashes one of the
world’s most deadly in cars equipped
with these systems they’re only enabled
45% of the time when they show up at a
dealership for service our work at MIT
suggests that false alarms confusion
over where and when and why these
systems are there to support you as part
of the cause we believe that education
is critical to understanding when and
where to use these technologies it
appears that many don’t realize that
with automation comes a responsibility
for increased education and systems to
promote proper use of the technology
I for one often get into my car with
backup cameras and remote sensing
backing up out of a parking spot whether
that’s parallel or perpendicular and
never fully inspecting the surroundings
around me I think many of us experienced
that today and we forget that this was
automation all created to assist us as
opposed to replace human sensing so
we’ve been hearing tales for some time
now of how automated vehicles are going
to enable safe more efficient
transportation however several recent
events including the tragic uber
fatality and several autopilot related
crashes illustrate that risks remain
factors related to these crashes suggest
that perhaps we need to be thinking more
about some of the softer elements of the
system in essence looking beyond the
core automation technology itself
technologists would say what’s missing
automakers tell us there are
the road to doing it better politicians
can we trust industry to do this for us
on the other hand us as drivers when’s
this driverless system going to be
available to take me to work
take my ailing parents to the doctor or
drop my kids off at school by itself
making my daily life a lot easier so
it’ll be more than fine tuning the
robots it’s gonna be more than building
a better car it’s going to be more than
moving the driver to the passenger’s
seat
our teen regulatory efforts will need to
be cast aside and as a society we’re
gonna need for an evolution versus a
revolution of the robots on our roads as
they will more likely be able to help
get you to safely to work
before they’re gonna be able to take
your parents that a doctor or your kids
to school on their own when we think
about these systems there are some tough
decisions to make because everything
around how we live and move may need to
be upended to get this right how safe is
safe enough just safer than the average
Joe or 50% safer more questions of who
to avoid and win a child playing in the
road or an older couple walking down the
sidewalk should we be thinking about
building automotive vehicles that aim to
replace human drivers it’s going to be
difficult to prevailed an artificial
system with the seemingly irrational
behavior of human drivers we often
ignore rules of the road speeding
squeezing through yellow lights or
making instantaneous movements based
upon almost unexplainable movements of
other cars around us when we think about
this how do we begin to develop systems
that we trust efficiently that we won’t
walk in front of it because we know
they’ll stop or we won’t merge in front
of because we know they’ll yield
expectations of the system are critical
to building trust and Trust is built on
experiences education may be a very
critical element in developing an
understanding of where the advantage of
automation are and where and when we may
neither take the wheel driver monitoring
you highlight in the news the thing that
we need to think about connecting to the
automation so it is aware of the drivers
attention to the world around us and
capable of forming the driver when
attention wanes
infrastructure improvements are
critically needed so that the automation
isn’t that spoofed by seeming failures
in our road infrastructure these and
other changes will not be easy and will
not occur without some stimulus we don’t
have a unified strategy for automating
driverless vehicles here in the United
States or much the rest of the
industrialized world federal policy
discussions are verily centered on what
I would consider is minor changes
today’s decades-old regulatory framework
that was created before microprocessors
and software ruled the world
without federal leadership we have 50
states heading in 50 different
directions clearly we need to develop
the foundations to enhance the system
and accelerate the process of saving
lives
politicians can make radical change
difficult in the timeframes needed to
keep up with this technology so I’d
argue that an alternative approach is
really needed automated vehicles may
seem radical and new but automation is
being leveraged by industries worldwide
in countless other settings there are
case studies from these industries that
we can leverage to build a foundation
for this safe mobility system that we
all desire for over 50 years the Federal
Aviation Administration has been working
to transform aviation system into the
trusted system we know today the
aviation industry has long known that is
the relationship between automation and
humans that is critical to safe flight
that’s why autopilot in particular is
licensed under the supervision of a
human operator the FAA has stressed
pilot education and training an
incredible investment in networked
infrastructure coordinated control and
tracking of aircraft data sharing and a
safety culture
oh this I see the safety culture aspect
perhaps being the most important the
automotive industry has long competed on
safety and perhaps a collaborative
safety culture is the critical element
to get organizations to collaboratively
work together to solve some of the most
difficult problems in education
infrastructure
and on a coordination of the automated
vehicles on our roadway the National
Transportation Safety Board is charged
with investigating major safety related
events related to the transportation
infrastructure like planes trains and
automobiles crashes so by default this
agency already has a role in
investigating the recent uber and Tesla
incidents but changes this an agency are
needed as well because cars by their
sheer number and the complexity of the
operating environment are going to crash
far more than planes and change trains
society can’t afford to wait a year for
an investigative report with from an
entity with the little enforcement power
when hundreds of thousands of automated
vehicles are on our roads and the
software systems fueling those are
changing by the day so a well-funded
nonpartisan safety focused scientific
entity really needs to be focused on
advancing the insights needed to
understand how to fuel the policies and
safety developments of the future if the
International Transportation Safety
Board remains as it is one day it will
not have the resources to respond to a
787 crash or the ability to impart
insights learnt fast enough to save
lives would you want one of your family
members or loved ones involved in a
crash as a result of a software
limitation that was known the Food and
Drug Administration has been regulating
the balance of benefits and costs and
medical device and pharmaceutical
discovery for decades with so many lives
lost to traffic collisions we need to
embrace technology such as automation
they can move the needle in a positive
safety direction while managing the
risks much likely manage the side
effects of a life-saving drug one has
the question should we be allowing way
moe to deploy tens of thousands of
automated vehicles on our roads without
some public understanding of the
cost-benefit risk model management plan
and tracking some system for side
effects that may occur after deployment
the clinical trial structure provides us
with a clear framework for testing
innovations in unknown
areas and stages we could test many
vehicle automation technologies in
similar way test the system components
check for integrated safety check for
efficacy on the road and collect more
data to garner the findings in short
requiring that companies demonstrate
they can walk before anybody is allowed
to run a strong post market to program
to ensure we have gotten this right or
making the changes needed maybe the most
important piece of the puzzle so I’m not
standing here and saying that we can’t
move fast I think we can a clinical
trial structure in the automotive
industry to the last days or weeks as
opposed to the years it takes in the
clinic is data coming from vehicles
comes much faster the key here is we
need to keep think carefully and
stepwise as a society before we allow
the technologists too much Liberty to
treat us as non-consenting test subjects
as such something new really may be
needed here if the government is going
to set stage by itself we need to think
big I can easily envision something as
transformative as a creation of a
department is in the context of Homeland
Security a new mission a new budget for
a new technology realistically this is
not easy our government does not move
fast nor too many of the rest of the
world nor is the technical skill needed
to understand this car technology really
there at the governmental level as
something we really need to think new
radical different perhaps in the context
of a private utility with technologists
and automakers leading the way but the
public sector charged with protecting
benefits and societal goods like safety
the context of working together in a
public-private partnership where
automakers and policy makers come
together to draft the policies that are
needed educational approaches testing
programs infrastructure changes ethical
dilemmas and answering questions such as
how safe is safe enough all under the
air of transparency affected has been
missing to date in essence trying to
create the alignment between stake hold
so that we can fuel the research and
efforts that are needed on the softer
side of this technology this is all
critical to building our trust in the
future of automated systems the status
quo is just not going to work when
robots are causing traffic jams
automated driverless vehicles are our
future but we can’t manage this
technology with dated measures that were
built before the era of software we know
that the free market left on its own is
not going to manage than weigh the cost
and benefits well by itself the biggest
problem here may be the will of all the
stakeholders to come together along the
pathway I described or an alternate
route to make the changes needed the
future of personal mobility and safety
is exceedingly bright when coupled with
the technology at hand and bumps on the
road that are going to occur like recent
incidents can be managed we’ve been here
before it was collaboration that got us
the moon and back safely politicians
wanted it companies wanted it
constituents wanted it so they made it
happen perhaps that’s the best framework
to be thinking about the modernization
and automation of our mobility ecosystem
in the long run automated vehicles will
enhance safety the question is can we
collaboratively work together to take
the reins of this public health crisis
and collaboratively strategically move
the safety faith fight forward I really
hope we can as too many people are dying
on the world’s roads every day and I’d
like my kids riding and driving in safer
roads it’s important we get this right
and do it quickly if we can automate it
in driverless vehicles will be the most
critical life-saving technology of the century
Please follow and like us: