{"id":30474,"date":"2023-10-25T09:05:51","date_gmt":"2023-10-25T08:05:51","guid":{"rendered":"http:\/\/www.spglobal.com\/mobility\/en\/research-analysis\/fuel-for-thought-waiting-for-autonomy.html"},"modified":"2023-10-25T09:05:51","modified_gmt":"2023-10-25T08:05:51","slug":"fuel-for-thought-waiting-for-autonomy","status":"publish","type":"post","link":"https:\/\/gamefootballmobileanimeiphone.com\/index.php\/2023\/10\/25\/fuel-for-thought-waiting-for-autonomy\/","title":{"rendered":"Fuel for Thought: Waiting for Autonomy"},"content":{"rendered":"

LISTEN TO A PODCAST ON THIS TOPIC
\nWITH S&amp;P GLOBAL MOBILITY EXPERTS<\/a><\/strong><\/p>\n

Five years ago, hype surrounding “autopilot” cars
\npeaked. But S&amp;P Global Mobility's updated forecast offers a
\nreality check of careful, limited adoption of self-driving
\ntechnology through 2035.<\/strong><\/h3>\n

A world of autonomous vehicles is likely to exist eventually,
\nbut for at least the next decade, that will be mostly in two very
\nspecific areas: geofenced robo-taxis operated by fleets in specific
\nareas and hands-off systems with various safeguards in personal
\nvehicles that require some form of driver engagement.<\/p>\n

As for a “World of Tomorrow” self-driving vehicle where you and
\nthe family can sleep in your seats and wake up 200 miles away at
\nGrandma's? The industry calls that “Level 5” autonomy, the highest
\nlevel of vehicular autonomy defined by the Society of Automotive
\nEngineers (SAE). That idea will stay firmly in the distance for the
\nforeseeable future, according to the latest forecast on vehicle
\nautonomy by S&amp;P Global Mobility.<\/p>\n

The ability for a consumer to buy a car that will drive itself
\neverywhere without the driver ready to take the wheel is unlikely
\nto happen by 2035. Level 5 implementation of fully autonomous
\ntechnology in the automotive retail fleet has zero representation
\nbefore 2035 in the S&amp;P Global Mobility forecast, “and probably
\nfor some time after that,” said Jeremy Carlson, associate director
\nfor the autonomy practice at S&amp;P Global Mobility.<\/p>\n

“Level 5 is essentially perfect replication of human driving,”
\nCarlson said. “A car that is able to drive anywhere and everywhere
\na human driver can go is simply not going to happen anytime
\nsoon.”<\/p>\n

Moreover, the S&amp;P Global Mobility forecast predicts a very
\nlow percentage of personal vehicles sold by 2035 will possess even
\na limited version of autonomous vehicle technology described as
\n“Level 4” under the SAE J3016 standard.<\/p>\n

The latest S&amp;P Global Mobility Autonomy Forecast details the
\nnumbers, timing, and countries for rollouts of varying degrees of
\nautonomous technology \u2014 based on a rigorous forecasting
\nmethodology analyzing
\nOEM and supplier strategies<\/a>, regulations, technology maturity
\nand cost, consumer desirability and willingness to pay, plus
\nextensive interviews both with automakers and across the supply
\nchain.<\/p>\n

\"\"<\/p>\n

This latest outlook continues to reflect the headwinds and
\nslower pace of development that both the automotive and tech
\nindustries have demonstrated over the past several years. But it
\nalso paints a stark contrast to the optimism of just five years
\nago, when the world was swept up in the excitement of a future of
\nautonomous vehicles.<\/p>\n

Now, S&amp;P Global Mobility presents a more realistic forecast
\nwhile also unveiling new data on the
\nintersection of autonomy and mobility-as-a-service (MaaS)<\/a>.<\/p>\n

Waypoints and indicators<\/strong><\/p>\n

As technology advances, of course, the array of automated driver
\nassist systems (ADAS) to compensate for driver inattention or error
\nwill proliferate \u2014 helping to reduce crashes, injuries, and
\ndeaths along the way. But it has proven incredibly difficult to
\ncombine those systems and sensors with the necessary predictive
\nsoftware and engineering into a vehicle that will carry us or our
\nloved ones to our destination unsupervised.<\/p>\n

There have been extreme instances of autonomous test vehicles
\nperforming as perfectly as a human would \u2014 perhaps the most
\nfamous early example came in 2015 when a Google test car navigated
\na situation where a woman in an electric wheelchair maneuvered
\nerratically on a residential street, while wielding a broom and
\nchasing a wayward duck.<\/p>\n

But serious malfunctions of the nascent technology have brought
\nmass-production \u2014
\nand public trust<\/a> \u2014 back to a more conservative reality. And
\nthe ability of autonomous systems to operate in the complex and
\nunpredictable world of interactions with human drivers (and
\nevolving driving circumstances) has proven a massive hurdle. From
\nthe endless cycle of software development and training of neural
\nnetworks, to continuous improvements in sensor hardware
\ncapabilities, to preparing for infinite corner cases that may
\narise, driving through the world has proven a complex task to
\nsolve.<\/p>\n

While there are growing numbers of driverless prototype vehicles
\nand pilot deployments already on city streets, S&amp;P Global
\nMobility forecasters do not expect that to become widespread and
\nbroadly accessible within the next decade. The vehicles you will
\nsee will be capable of driving themselves in MaaS use cases. You
\ncan think of them as “robo-taxis,” replacing anything from the NYC
\nyellow cab to rideshare services such as Uber and Lyft.<\/p>\n

Robo-taxis come before personal vehicles<\/strong><\/p>\n

Robo-taxis are expected to be carefully geofenced for the
\nforeseeable future \u2014 offering revenue service only within
\nwell-mapped areas where they have already been extensively tested,
\nCarlson predicts. That's the model for current services from
\nCruise, Waymo, Motional, Didi, Baidu and others. Waymo was first to
\nlaunch a commercial service in certain Arizona suburbs and is
\ntesting in several major California cities. Cruise began testing in
\nSan Francisco in late 2020 (and operating limited public robo-taxis
\nin 2022) and is now expanding to its seventh city. Several Chinese
\nmobility providers \u2014 from Baidu to Didi to Pony AI and more
\n\u2014 have similar approvals for testing and early, limited MaaS
\noperations in various cities and regions as well.<\/p>\n

The cost of the multiple camera, radar and lidar sensors and the
\nhigh-performance computer systems, chips, and semiconductors to
\nenable the technology will remain sufficiently high that most
\npersonal-vehicle buyers simply won't pay for the functionality
\n\u2014 but taxi and ride-sharing services may. Those businesses want
\nto keep their vehicles on the road 20 or more hours a day, stopping
\nonly to refuel or recharge the batteries. With that amount of
\nuptime, they can earn revenue without the inconvenience of drivers,
\nwho need to be paid, change shifts, and often stop for breaks. All
\nof those factors cut into potential revenue hours; robo-taxis have
\nnone of those constraints.<\/p>\n

The ultimate goal \u2014 and the vision of lots of movies and
\nother science fiction over the past 75 years \u2014 is an unpiloted
\nvehicle where a rider can get in, input where they want to go, and
\nrelax in the back seat. No matter the destination (anywhere on
\nearth a vehicle can go), the weather, or road conditions, the
\nvehicle simply drives there itself. “The industry is a long way off
\nfrom that,” Carlson says. “But forms of Level 4 are operating today
\nin limited numbers and have a promising decade of development and
\ngrowth ahead.”<\/p>\n

A decade of tech hype<\/strong><\/p>\n

Fifteen years ago, the seminal DARPA Urban Challenge proved
\nvehicles could navigate themselves to a destination through unknown
\nand unpredictable environments. What followed was a wave of hype
\nsplashing over what many saw as the imminent arrival of
\nautonomous-driving technology across the automotive spectrum.
\nBillions of dollars in capital expenditures were invested into
\nautonomous vehicle development, with substantial progress but
\nvarying degrees of success. Bold visions have been espoused, and
\nwhile a few waypoints have been reached, the grandiose
\nproclamations have not been achieved.<\/p>\n

Part of the overoptimism, Carlson says, came from technology
\nfirms being the first to promote driverless cars in a big way. From
\nthe multiple lidar firms commercialized after the DARPA event, to
\nGoogle, Uber, and Tesla, self-driving tech was assumed to be
\nlargely a problem of sensors and software \u2014 and software could
\nbe developed fast, right? The peak of autonomy hype came in 2018 at
\nvenues like the Consumer Electronics Show.<\/p>\n

Automakers and tech companies (and more than a few consultancies
\nand industry think tanks) were swept up in the fervor, with
\nexecutives and investors catching the fever \u2014 despite
\nprotestations from engineers that even limited autonomy was a
\nreally, really hard problem.<\/p>\n

It was at CES in 2020, just before the global COVID-19 pandemic
\nshut down much of business as usual, that it became apparent lidar
\nmakers would need more time to meet the cost and performance goals
\ntheir customers \u2014 the automakers \u2014 had set, pushing back
\ntimelines for introductions of automated driving systems in Level
\n2+ and Level 3. While lidar is a key enabler of autonomous
\nvehicles, a similar but also unrelated reset of expectations
\noccurred with autonomous vehicles themselves.<\/p>\n

Now, S&amp;P Global Mobility states, the line of thinking is
\nmore grounded: Unbridled optimism has given way to the slow,
\nmeasured development of new, highly complex technologies. Tech
\ncompanies must fuse reams of digital input from a variety of
\nsensors, model the results, and pass that model through software
\nthat determines what behaviors to require of the vehicle as a
\nresult.<\/p>\n

Obstacles to progress<\/strong><\/p>\n

To understand how monumentally difficult this accomplishment
\nwill be, an examination of each step in the route to fully
\nautonomous technology is required. And each step is steeper \u2014
\nand more costly and legally binding \u2014 than the one before.<\/p>\n

New-car shoppers today will have found most of the latest
\nvehicles now offer a variety of ADAS technologies, as either
\nstandard or optional equipment. The lowest tiers of the SAE
\nautonomy scale, Level 0 and Level 1, are used for functions a
\ndriver can invoke to control individual aspects of a vehicle's
\ntravel automatically, or otherwise assist the driver with alerts or
\nspecific emergency interventions. From adaptive cruise control to
\nlane-departure correction, these systems compensate for driver
\nerror or inattention, but generally only in one direction of
\ntravel: either longitudinally (headed down to the path of travel)
\nor laterally (side to side within a lane).<\/p>\n

Integrating several Level 0 and Level 1 features allows the car
\nto effectively drive itself within lanes (usually on limited-access
\nhighways), which can extend the vehicle into Level 2. S&amp;P
\nGlobal Mobility sees a huge proliferation of Level 2 systems \u2014
\nbecoming the de facto standard equipment for many vehicles \u2014
\nalong with a more capable subset of what has come to be known as
\n“Level 2+.” Standard Level 2 can be found on vehicles from the
\nlikes of Tesla, Toyota, and Honda, but the driver must prove they
\nare paying attention by keeping hands on the steering wheel \u2014
\nas the vehicle will only provide emergency avoidance support (like
\nbraking) but is not capable of responding to more complex changes
\nin traffic conditions.<\/p>\n

The system tests for driver interaction by searching for
\nconstant small inputs into the torque sensor of an electric power
\nsteering system, which correspond to natural human movements on the
\nwheel. The problem is that this check process is easily defeated,
\nas dozens of YouTube videos show: Tape a half-full water bottle to
\nthe steering wheel, and road imperfections cause the water to
\nslosh, jostling the wheel enough to “prove” a human has hands on
\nthe wheel and is, presumably, paying attention. Capacitive sensors
\nlike those of smartphone or touchscreen displays provide more
\nintelligence but require hands-on engagement.<\/p>\n

The only way to provide a true “hands-off” system \u2014 while
\nstill requiring the driver to keep eyes on the road, as opposed to
\ntheir attention on something else \u2014 is eye-tracking cameras.
\nThose are now built into a handful of models from a few makes but
\ngrowing rapidly and with strong upside over the next decade. The
\nresulting hands-off driving systems available today with driver
\nmonitoring systems are known as Super Cruise (General Motors),
\nAutopilot (Tesla), BlueCruise (Ford) and various others.<\/p>\n

European regulations will mandate the installation of these
\ndriver monitoring cameras in mid-2026. It's a stiff standard, but
\nthe only one that effectively guarantees that while a driver's
\nhands may be off the wheel, their eyes are still on the road ahead
\n\u2014 meaning they can retake control of the car in an emergency
\nwithin a second or two, rather than the five to 15 seconds (or
\nmore) that's required if hands, eyes and mind are distracted by
\nsomething else entirely.<\/p>\n

The combination of Level 2 automated driving with driver
\nmonitoring pushes into Level 2+ and is the area of focus for nearly
\nevery automaker in the industry today. It includes a very capable
\nform of sensing and automated control while ensuring the driver
\nremains engaged and capable of supervising its operation. As
\nimportant as any part of the design, it also maintains liability on
\nthe driver's side of the equation.<\/p>\n

Then there's Level 3. This is autonomous driving by a vehicle
\nwithout the driver needing to pay attention, but only in strictly
\nlimited circumstances. “Drivers are no longer required to
\nsupervise,” Carlson explains. “Instead, the vehicle is responsible
\nfor all of the object detection, response, and actuation of
\nappropriate control responses.”<\/p>\n

This is a huge leap, both technologically and legally. The
\nsignificant difference is that liability for any crash caused by
\nthe Level 3 (or above) vehicle operating in autonomous mode shifts
\nfrom the driver to the automaker, as the driver isn't required to
\nbe constantly engaged in the driving (or supervising) task. That
\nmeans automakers will be exceptionally careful in taking this step,
\nCarlson predicts. As such, S&amp;P Global Mobility forecasts far
\nless Level 3 penetration by 2035 compared with the pair of Level 2
\ntiers.<\/p>\n

Because robo-taxis are limited to certain operating areas and
\nconditions, they fall into what's called Level 4. Carlson expects
\nconsiderably more services with potentially infinite expansions of
\nthose geofences. Fleet operators see a potential business model
\nunder which the added costs of Level 4 technology could be offset
\nby greater uptime over a vehicle's lifetime. But automakers do not
\nsee sufficient willingness among consumers to pay an additional
\nfive figures for Level 4 functions in mass-produced vehicles over
\nthe next 10 years.<\/p>\n

\"\"<\/p>\n

Where do we stand today?<\/strong><\/p>\n

Tesla's much-hyped Autopilot system to the contrary,
\nMercedes-Benz has been at the forefront most recently as the
\nindustry pushes into Level 3 and beyond. Its Drive Pilot system in
\nGermany relieves the driver of having to pay attention while the
\ncar drives on limited access roads \u2014 at less than 60 kilometers
\nper hour (36 mph). This low-speed operation is the opposite of the
\nLevel 2+ approach taken by GM and others, which started on
\nlimited-access highways only and is now expanding to certain
\ndivided secondary roads.<\/p>\n

In Europe, regulatory updates are underway to boost the upper
\nlimit to 130 km\/h (80 mph), recognizing that automated highway
\ndriving may save more lives through crashes averted than in
\nlower-speed urban environments \u2014 but also enabling automakers
\nto deliver more real-world value to potential buyers. Still,
\ncarmakers are taking it slowly and steadily. The rollout of
\nautonomous driving features has proceeded more on traditional
\nautomaker schedules for developing new features, rather than in
\n“tech time” where features are conceived, prototype, launched and
\nrevised on the fly.<\/p>\n

But this tech industry influence is still the central theme of
\nthe movement toward the software-defined vehicle, one that utilizes
\nover-the-air (OTA) updates to maintain and enhance functionality
\neven after the point of sale.<\/p>\n

MaaS robo-taxis, for now<\/strong><\/p>\n

The S&amp;P Global Mobility forecast predicts fewer than 6% of
\nlight vehicles sold in 2035 will have any Level 4 functionality, as
\ndescribed by the SAE J3016 classification. Early Level 4
\nimplementations in personally owned vehicles offer advanced parking
\nfunctions, often with the support of infrastructure. But many
\ntechnology providers remain focused on the long-term potential of
\nscaling autonomous vehicles in fleets supporting MaaS business
\nmodels.<\/p>\n

There are positive examples of autonomous vehicles performing as
\nwell as humans in today's pilot programs in places like San
\nFrancisco and Phoenix in the US, and Beijing, Shanghai, and
\nGuangzhou in mainland China. But these same vehicles can still be
\nconfounded by complex traffic scenarios the next minute or the next
\nday, giving regulators and consumers alike reason to be
\ncautious.<\/p>\n

Mobility-as-a-Service (MaaS) and robo-taxis are nonetheless
\nexpected to lead the transition to an autonomous vehicle future,
\neven with the relatively cautious growth ahead. There are growing
\nnumbers of small-scale deployments in certain cities around the
\nworld. But S&amp;P Global Mobility forecasters do not expect that
\nto become widespread and broadly accessible within the next
\ndecade.<\/p>\n

MaaS-equipped vehicles and robo-taxi applications are expected
\nto represent less than 800,000 vehicles sold globally in 2035.
\nRobo-taxis will be carefully geofenced for the foreseeable future
\n\u2014 offering revenue service only within specific areas where
\nthey have already been extensively tested, Carlson predicts. But
\ntheir high rate of utilization can be nonetheless effective at
\nbringing new mobility options to some consumers and new revenue
\nstreams to automakers and mobility providers.<\/p>\n

\"\"<\/p>\n

Owen Chen, senior principal analyst from S&amp;P Global
\nMobility, explains that robo-taxi development and commercialization
\nis a complex and multi-stage process, which can be summarized in
\nthree stages:<\/p>\n