AUTOMOTIVE MULTIMEDIA AND COMMUNICATIONS

Detailed system and semiconductor demand analysis for in-vehicle infotainment, telematics and vehicle-device connectivity features.

February 10, 2015 06:47 rlanctot

General Motors’ OnStar vehicle connectivity solution was once shared under license with Audi, Acura, Isuzu, Subaru, Lexus and Volkswagen – an offer which was later terminated. OnStar made a second (public) attempt at licensing its technology to competing auto makers in 2009, but subsequently pulled that offer back as well.

In view of the vehicle security and privacy crisis facing the auto industry, it may be time for OnStar to reconsider the licensing option - that is, if OnStar can firmly establish its security credentials. Outside of OnStar, the auto industry collectively stands like a deer in the headlights as reports of vehicle vulnerabilities multiply without a clear solution in sight.

But OnStar - itself - ended up in the headlights of the television program '60 Minutes' when a thinly disguised GM vehicle was used to demonstrate the ease with which vehicles may be hacked.

In retrospect, there are those that view OnStar's reversal of its licensing decision as a major mistake, a lost opportunity to dominate the connected car industry globally. Licensing will pose fundamental challenges to competing car companies and the architectural choices that they have made governing their connectivity systems along with the partnerships they have forged.

Driving a licensing strategy, though, is the global industry demand for an immediate solution to ongoing and embarrassing revelations regarding vehicle hacking. And security shortcomings will short circuit plans for self-driving cars.  So much is at stake.

In the wake of ADAC’s hack of BMW vehicles in Germany, it is clear that the challenge of vehicle security must be solved first before governments can comfortably welcome fleets of autonomous vehicles on public roadways. But yesterday’s release of U.S. Senator Edward Markey’s report on vehicle security and privacy concerns and Sunday’s ‘60 Minutes’ episode on “DARPA Dan” Kaufman reveals the U.S. government stepping into a leadership role on both fronts.

http://tinyurl.com/ncfelr6 - Markey Report Reveals Automobile Security and Privacy Vulnerabilities
http://tinyurl.com/p8429cq - 60 Minutes Episode on ‘DARPA Dan’ Kaufman

Markey’s report highlights the lack of coherent and consistent privacy policies in the automotive industry – in the context of an industry that is culling massive amounts of data from its vehicles and enabling unfettered access to increasing volumes of that data to third parties. But of greater concern is the porous state of vehicle security, where the industry has either chosen to hide its head in the sand or, worse, confidently claimed to have a handle on the problem without actually recognizing its scope – Exhibit A: BMW and, now, GM.

Most interesting is that in the ‘60 Minutes’ episode focused on the Defense Advanced Research Projects Agency’s efforts to combat hackers attacking government and private agencies, CBS chose to highlight what some might regard as the lowest priority target – the automobile. Nevertheless, the hacked car has clearly become the most popular point of vulnerability to use when trying to rouse the concern of the general public.

Both the Markey report and the ‘60 Minutes’ episode point to the following conclusion:

  1. We must resolve the issue of vehicle security before tackling autonomous driving
  2. There is a role for the government, but that role likely lies in DARPA – or maybe NASA – NOT NHTSA (U.S. National Highway Traffic Safety Administration)
  3. Vehicle connectivity is the solution, not the problem

Vehicle security virtually mandates vehicle connectivity. Wireless vehicle connections are necessary to properly secure vehicles and enable vehicles to preserve that security with the ability to receive updates of prophylactic code. Wireless connections are also necessary for vehicles to be able to communicate when intrusions have occurred.

The Markey report and the ‘60 Minutes’ episode also highlight how far ahead of the entire industry OnStar remains – 18 years after its launch. Only OnStar (and Porsche) have the ability to enable the remote slowdown and immobilization of one of its cars in the event of a theft. This feature now stands out in the context of the security vulnerability as a default safety response.

OnStar, in addition to Tesla, also stands out for its perfection of software updates – a pre-requisite for ensuring up-to-date on-board software code. In effect, OnStar’s fundamental architecture anticipated and provided for secure connectivity across the entire GM line up.

This is not to say that OnStar hasn’t doubled down on its existing security solution.  Indications from the Renaissance Center, GM's headquarters, is the the company is hard at work hardening its platform in anticipation of the launch of its next generation.  The next step for OnStar and, in fact, the entire industry, is to enable the telematics control unit to act as or be connected to a central vehicle hub or “brain” to monitor all systems on the vehicle network for intrusions.

Here it is worth noting that the IOactive vision of segregated vehicle systems as recommended by that company in 2014 to enhance security, runs counter to the current industry shift toward integration.

It remains to be seen whether OnStar, like Tesla, will choose to share or license its security secrets with the industry.  In fact, there is a wide open opportunity for any of a dozen entities to bring a solution to market.  But OnStar maintains a technological edge after 18 years and has shown previous interest in licensing its technology.

It’s worth noting that, in the end, the core value proposition that ultimately distinguishes OnStar might not be Wi-Fi or apps. It may yet be the safety and security that have always been hallmarks of the OnStar brand.

*This blog was revised @ noon EST Feb. 10 to reflect more accurate information regarding the current state of OnStar vulnerability.


February 5, 2015 13:05 rlanctot

It would be funny if it weren't so serious, shocking if it weren't so completely unsurprising. The German automobile club, ADAC, discovered a simple vulnerability in BMW's ConnectedDrive system that exposed vehicle data and controls. BMW and ADAC cooperatively announced last week that the problem had been corrected.

BMW had failed to implement even the most basic security measure - HTTPS encryption - thereby enabling ADAC to access BMW vehicles with embedded SIMs associated with ConnectedDrive. The access to the SIM was anabled by creating a fake wireless network to which the cars attempted to connect. The fix was made in BMW's back-end system thereby avoiding any need for an update of any software code in its vehicles.

You could call what ADAC did a "hack" or a stunt. It's an approach used by law enforcement and criminals around the world and sufficiently common to be alarming.

(There are those who suggest that the ADAC hack was more complex and that the BMW security shortcomings were more severe - some of which may yet to be resolved. Details here: http://tinyurl.com/o4z3nzs)

ADAC detailed the affected vehicles and recommended broader testing of all cars and proper certification for all car makers. No such required testing or certification exists.

"ADAC demands state-of-the-art protection of in-car computer technology against manipulation and illegal access," noted an ADAC representative. "Such protection must be based on standards long since operative in other industries
(e.g. IT industry).

"Moreover, said protection needs to be confirmed by an impartial body, e.g. via Common Criteria certification through the Federal Office for Information Security (BSI) in Bonn, Germany – or related organizations in other countries (refer to www.commoncriteriaportal.org)."

Why ADAC?

Perhaps the most interesting aspect of the entire affair is that ADAC was probing BMW's for diagnostic data identifying scheduled or unscheduled service or repair opportunities. "ADAC commissioned an external expert to analyse the information which vehicles transmit to the manufacturer via BMW Connected Drive when an inspection or repair is due. The objective was to determine whether independent workshops might be at a disadvantage and whether ADAC should step in to protect consumer interests."

So, ADAC discovered the vulnerability in the process of trying to preserve the right to repair privileges of independent repair shops. ConnectedDrive and other such embedded connectivity systems from car makers are perceived as giving OEM dealers an unfair advantage in servicing their cars.

Says ADAC: "Although this was never intended, the investigations revealed security loopholes, prompting the publication of the findings."

BMW was fairly blase about the matter. My contacts at BMW brushed it off as a well-known problem that amounted to nothing more than an oversight. The greatest risk posed by the vulnerability was that it made the cars easier to steal.

There are a few key takeaways to this incident:

  1. How did BMW miss this vulnerability?
  2. How come BMW's suppliers failed to uncover this vulnerability?
  3. How and when will governments take on the responsibility for certifying vehicle security?
  4. Shouldn't we resolve the certification of vehicle security before we allow autonomous vehicles onto public roadways?
  5. What are the implications for the broader universe of connected things?
  6. Are cars a special case requiring a level of certification that is not relevant to other devices?
  7. Is ADAC (and by extension all auto clubs in markets around the world) getting into the vehicle security certification business? Is ADAC a whistle-blower? Or is ADAC intending to extend its testing to all connected cars?

The ADAC hack definitely has a your-fly-is-open quality to it. But the implications are serious.

Car makers have repeatedly demonstrated their inability to comprehend the nature and scope of the vehicle security issue. Companies such as Red Bend, Blackberry and OpenSynergy have solutions and those solutions are seeing wider application.

But the scope of the problem is pervasive touching multiple systems and points of access on cars requiring multiple layers of protection. In and of itself this suggests that certifying security will be no easy process. Nevertheless, BMW's open-fly problem reveals a massive industry blindspot that must be corrected.

OnStar Hacking Update:

I have to correct an impression I left in a previous blog that GM's OnStar had never been hacked. OnStar was hacked about five years ago and has since corrected its vulnerability. Details:

http://www.nytimes.com/2011/03/10/business/10hack.html?_r=1&

http://www.autosec.org/pubs/cars-usenixsec2011.pdf

http://www.eweek.com/c/a/Security/GMs-OnStar-Ford-Sync-MP3-Bluetooth-Possible-Attack-Vectors-for-Cars-420601


January 26, 2015 07:37 rlanctot

Over the years I have been frequently asked at automotive events why car makers don’t all agree on a single user interface for the car stereo and be done with it. It’s a good question, especially now that it is sometimes almost impossible to FIND the car stereo in some cars.

But the question has changed over the years and the more up-to-date version of the question is: Why don’t car makers give me a place to dock my tablet computer (or other mobile device) in the car and be done with it? This is also a good and logical question.

What lies behind these questions is the fact that user interfaces in cars, for car stereos or whatever is happening in the console/centerstack area, are entering a new phase influenced by regional preferences, new display technologies, government mandates and mobile devices. The bottom line: Touch screens are coming to cars. It just so happens that a large chunk of the German automotive industry is out of touch with this trend.

For years, about 13 to be exact, the German auto industry has operated under the HMI (human machine interface) thrall of BMW’s i-Drive hardware controller. The i-Drive controller – a rotating knob located in the front seat center console – allows the driver to do everything from changing radio stations to entering destinations one character at a time.

Introduced on the BMW 7 Series in 2001, the i-Drive is universally loathed and loved. Most auto enthusiasts despise the i-Drive, while those less than religiously devoted to the art of driving find it amusing and convenient.

BMW is entitled to kudos for cleverness, but it’s time to retire this relic once and for all. The i-Drive and its equivalents have become anachronisms in a post-iPhone world.

Daimler and Audi (and Acura, Infiniti and Lexus) all followed BMW’s i-Drive lead. All of these companies are now faced with a major HMI rethink as hardware controllers are increasingly seen as contributing to eyes-off-road-time rather than mitigating driver distraction.

Touch screens have emerged as the preferred alternative throughout the world, particularly in Asian markets, motivated by the widespread adoption and use of mobile devices with touch screens. Hardware controllers in cars suddenly look about as novel and clever as Atari joysticks.

Now, don’t get me wrong, these devices are awfully clever – and I do mean awfully. Having sat through multiple presentations regarding the efficacy of hardware controllers (and touchpads) I can confirm that it is quite possible for two groups of similarly trained engineers to draw completely different conclusions from the same research. Or it is equally possible for those two groups to conduct their research in such a manner as to produce a predetermined outcome. It happens.

German auto makers have managed to avoid touch screens based on the perception that drivers would need to change their focal length to look at and touch the display. More than one automotive engineer told me that they did not want their customers to smudge the screen with fingerprints.

They certainly could not argue for the lower cost of the touch screen because the hardware controller added both hardware AND software cost to the development of the system. With car makers seeking to rapidly ramp up in-vehicle app integration platforms, hardware controllers are introducing excess cost and complexity.

But the commitment of these German car makers to non-touchable screens dictated that the screens literally be placed beyond the reach of the driver. So these car makers must now yank out the hardware controllers, reconfigure their software and move their displays within reach of the driver. (It may also mean the demise of pop-up or peekaboo displays that appear on vehicle ignition. Yes, that means you Audi et. al.)

So today, in a world increasingly dominated by touch screens, speech and gesture recognition, eye tracking and steering wheel controls, the hardware controller has got to go. It has outlived its usefulness.

Quaint and clever though they may be, hardware controllers are out and touch screens are in. More or less immune from this rethink are North American auto makers which mostly ignored the hardware controller bandwagon. U.S. car makers are hereby rewarded for their slow pace of technology adoption.

Sometimes it’s good to follow slowly – maybe the driver in front of you doesn’t know where he or she is going. Speaking of following slowly, Volkswagen was slow to adopt hardware controllers with the rest of its German industry brethren thereby enabling a more rapid shift to touch displays. The company was the first to break ranks with the joystick crowd. It will be interesting to observe how and when and at what cost the rest of the German auto making community gets in touch with this trend.


January 14, 2015 07:00 rlanctot

The auto industry is abuzz over Google’s plans to bring Android to automobile dashboards as a native operating system with all that that might imply or enable. This is not to be confused with Android Auto, the smartphone connection proposition shown by several car makers and suppliers at CES in Las Vegas last week. The next phase, for Google, is having its OS act as the native operating environment in the car.

Unlike Google, Blackberry has been playing the automotive OS game for a couple decades with its QNX real-time operating system. Suddenly the nifty black magic that QNX has been able to pull off – delivering an operating system capable of seamlessly keeping pace with consumer electronics trends and updates without requiring hardware changes – is beginning to get some appreciation – from Ford among a host of other car makers.

That old Blackberry black magic is highlighted in the Panasonic suite at the North American International Auto Show in Detroit this week. Not only is Panasonic showing early versions of Ford’s upcoming SYNC 3 head units using Blackberry’s OS, the company is also privately showing its car company customers and prospective customers an upgradable dashboard concept – along with a few other clever solutions and systems for hands-free app and smartphone integration including Aupeo!

Consumers readily understand software updates, which impact everything from smartphones to TVs to PCs. Hardware updates are usually only associated with desktop or notebook computers.

Many owners of Apple iPhones, though, were forced by the latest iOS update to delete some of their apps to make room for the update. This is precisely the challenge facing car makers interested in implementing Android as the native automotive OS. Only in the case of Android, the issue is more severe.

Keeping pace with Android updates and upgrades usually or eventually means getting an entirely new device, because the old device is incapable of supporting the processing or storage demands of the new Android version.

Android knows no bounds.

The significance of the upgradable Panasonic system is that it will allow customers to upgrade and replace the more fungible elements of the car stereo system in a single dock-able module with a proprietary hardware interface. This is precisely what adopting Android in cars will require.

Panasonic, though, is not spinning the modular update as an Android in the dashboard story. Panasonic’s official statement:

“Today’s consumers are expecting constant upgradability – something they’ve been taught based on upgrading their consumer electronics. Their expectation is that this functionality is available across all their electronics, including their vehicles. This is a growing challenge for auto manufacturers due to the different lifecycles of consumer products, specifically smartphones and vehicles. Panasonic has realized this problem, and has created a future-proofing solution that offers more.

“Panasonic Automotive is not yet releasing details on its modular embedded concept. However, this consumer-driven technology is being developed as a removable, upgradable system so consumers can update to the very latest software and hardware. This not only expands the user experience through software, but, with upgradable hardware, it becomes easy to support any future technologies and devices brought into the vehicle – a differentiator for the Panasonic system.”

Panasonic is not disclosing details regarding the modular concept, but its content could include the Bluetooth, USB and Wi-Fi elements involved in connecting to the driver’s smartphone along with the storage, processor, media support and maybe even the embedded OnStar-like wireless connection. The Panasonic demonstration in Detroit follows public and private announcements from nVidia and Qualcomm at CES regarding their plans to encourage and enable upgradability in their automotive offerings.

Automotive announcements at CES were full of new firmware over-the-air update propositions – a recognition of the explosive proliferation of automotive software for everything from safety systems to infotainment. But hardware updating is something new, different and challenging.

Software updates provide a means to preserve or enhance vehicle value after the sale of the car and can include everything from new apps or map updates to modifications for safety system and airbag algorithms. Software updates can be brought to the car by the dealer, a customer’s smartphone, an in-dash USB port, Wi-Fi or via a high-speed data port connection.

Software updates are normally free and are provided as either a customer service or as a necessary means to preserve the functionality of on-board systems. Hardware updates, meanwhile, are generally not available due to the logistical challenge and cost. If the customer needs or wants a hardware upgrade, it’s normally either time to get a new car or find an independent installer to replace the existing system.

Panasonic is suggesting the creation of an aftermarket for hardware updates intended to keep automotive hardware fresh and up-to-date throughout the 11-year+ life of the vehicle. But car makers never had to seriously contemplate this prospect while using Blackberry’s QNX OS or even Linux. But the arrival of Android has introduced the need.

Hardware updates will be necessary to support new versions of Android. It is no coincidence that CloudCar, an Android advocate within the supplier community, has itself shown a GPU upgrade solution. CloudCar has yet to find any takers for its nVidia-based vision for auto system upgrades.

Another supplier, Cybercom, has proposed its “Infotainment-on-a-stick” concept modeled after Google’s Chromecast HDMI plug-in device. So far, the Cybercom initiative remains only a concept.

Does this mean we will bring our cars in to dealers in the future for freshening up of the hardware? Or does it mean that we will be able to upgrade a car with low-end infotainment performance to turbo-tainment a year or two after the original purchase? Does it mean that the entire infotainment package can and will be sold separately? Is hardware upgradability a precursor to Android’s entry into dashboard systems? The answers to these questions reside in the minds of car makers visiting Panasonic this week.

Cybercom Infotainment-on-a-Stick


January 7, 2015 13:12 rlanctot

President Obama’s appointment of Mark Rosekind to lead the National Highway Traffic Safety Administration (NHTSA) was a wake-up call to the transportation industry...and Congress. Following closely on the heels of Congress deciding to suspend year-old rules for truck drivers requiring two nights of sleep before a work week, the appointment of Rosekind – an expert in pilot fatigue – was an incisive way for Obama to demonstrate, yet again, that he is anything but a lame, lame duck.

The decision by Congress ignored requests from Transportation Secretary Anthony Foxx to preserve the limits on drivers. With the lapsing of the year-old rules, thanks to Congress, drivers will be allowed to work as many as 82 hours over an eight-day period.

NHTSA estimates that drowsy driving for all kinds of drivers and vehicles in the U.S. causes more than 100,000 crashes a year, resulting in 40,000 injuries and 1,550 deaths. The fatality rate is half the estimate for the number of distracted driving fatalities, but NHTSA says drowsy driving is underreported.

Nearly 4,000 people are killed annually in accidents involving large trucks, according to NHTSA. While the overall number and rate of fatalities declined in 2013 including the figures for incidents at intersections and involving alcohol-impaired drivers, the number of fatalities involving large vehicle collisions increased 0.5%.

NHTSA has conducted a massive amount of research on driver fatigue and it is clear that this research has immediate application in the large vehicle market. (NHTSA’s drowsy driver research can be found here - http://tinyurl.com/nsoj66x.) With Rosekind’s leadership perhaps NHTSA can not only advance the understanding and mitigation of drowsy driving in the large vehicle market but also generate insights with application in the wider consumer market.

Companies like Seeing Machines that are working mainly in the commercial market today, stand to benefit from the introduction of fatigue detection and mitigation systems for consumers. Seeing Machines is already working with GM and Takata on just such a system. This is also likely to be a rich target environment for app and device-based companies such as Anti Sleep Pilot.

In the end, Rosekind's appointment likely points the way to the wider deployment of cameras and sensors inside vehicles for identifying drivers, recognizing gestures and assessing driver attentiveness. The Rosekind appointment could well point the way toward new life-saving research and product development guidance from NHTSA.


January 4, 2015 12:46 rlanctot

What if you were a car maker and I told you I had an app that would work with existing hardware in a car to deliver – at no cost – local (and national and international) news, weather, sports and traffic information along with a wide range of curated music and spoken word content? Sounds incredible, right?

What if I told you the app would also provide emergency alerts and would use an existing network with ubiquitous coverage that would work in your car, your home, at work and online? And what if that app were capable of delivering detailed traffic reports including information regarding the causes of backups and the status of the local response?

What if I told you the app would work simply with existing controls in the car without distracting the driver?

That is exactly what the existing car radio is today: a free ad-supported application with location-relevant content and services.

But some press reports in 2014 suggested that the AM/FM radio’s grip on the dashboard was more tenuous than many realized. GM’s decision to remove HD Radio from five 2015 model-year vehicles – the Chevy Traverse, Silverado and Impala and the Buick Enclave and Regal – had observers proclaiming the premature demise of HD Radio. This report came on the heels of BMW’s removal of the AM band from the BMW i3 due to interference from the electric motor.

The radio at risk

Many of the competing alternatives to the car radio are working hard to replicate all or part of the contextual AM/FM content delivery experience via a cellular IP connection. Those alternatives include but are not limited to Aupeo, TuneIn Radio, Aha Radio, Stitcher, Rivet Radio, iTunes Radio and OmnyApp to say nothing of the streaming audio contenders Pandora, Beats, Rdio, Amazon Cloud, Rhapsody, Spotify, Deezer, Douban, and a growing roster of competitors around the world.

Cars may be losing their CD players and traditional rearview and side mirrors, but cars will definitely have radios in 2015. The only problem is that you may not be able to find the radio in your new or rental car.

The growing variety of in-dash systems and smartphone integrations are making the traditional AM/FM radio hard or harder to find. A study released earlier this year by Digital Radio UK intended to assess the usability of digital radio and the future for radio in connected cars found:

  • 30% of the cars (in the study) had a system to integrate the smartphone with the dashboard
  • 100% of manufacturers have developed their own proprietary connected system
  • 100% of manufacturers are working with Apple on CarPlay, Google on Android Auto, or both
  • 50% of manufacturers are working with MirrorLink
  • Internet radio can have up to eight (8) menu layers to find a station
  • Internet radio suffered from poor coverage and buffering
  • In connected cars “Radio” will be used to describe audio services that aren’t broadcast radio

The Digital Radio UK study concluded:

  • Consumers could find it harder to find broadcast radio
  • Internet radio can be complex to use and could cause driver distraction
  • Broadcasters and vehicle manufacturers should develop a new radio experience for digital dashboards

The study – The Digital Dashboard Audit, November 2014 – offers a five-point action plan for car makers to coordinate their digital dashboard development efforts. Of course, the issue in the UK goes beyond the analog AM/FM experience.

(The Digital Dashboard Audit contrasts with competing reports regarding the use of LTE wireless networks to deliver radio content. These other studies – Teracom in Sweden and other companies – envision a complete switchover to wireless delivery from broadcast with all the related costs and burden on the LTE network. An example can be found here: http://tinyurl.com/q6pqzlc - “LTE Broadcast – The Next Hyped Broadcasting Challenger”)

Digital radio to the rescue

The UK is in the forefront of a pan-European effort to adopt digital radio. Norway, Denmark, Switzerland and the UK are the furthest along in this effort, which is expected to ultimately include a shutdown of the analog network in some countries.

The UK is closely monitoring adoption of and the amount of listening time devoted to DAB+, the prevalent digital standard in use in Europe. According to data from CAP Automotive Ltd., the monthly car registration volumes by digital radio option status and availability show 60% of cars registered in the UK in Q3 of 2014 with standard fitment of digital radio technology.

Germany and The Netherlands have launched DAB+ services with eight other European countries launching regional DAB+ broadcasts including France, Italy and Sweden. Australia is also a DAB+ leader. This is in contrast to the U.S. where HD Radio technology has been widely adopted by all car manufacturers, 2,200 broadcasters with more than 1,500 multicast stations transmitting to the more than 19M receivers that have so far shipped.

Digital radio is arriving globally just as IP-delivered content via connected smartphone apps are hitting the market. The challenge for car makers is twofold: distraction and confusion. The temptation for some car makers may be to throw the AM/FM radio out altogether.

Analog AM/FM is facing challenges even under the best of circumstances without an existential threat from IP-delivered options. The quality of the AM signal in the U.S. continues to deteriorate due to increasing interference from other signal emitters, such as power lines, and some listeners complain that even analog FM reception quality has deteriorated.

Any radio station in a storm

The deterioration of radio broadcast signal quality is significant because radio is the original emergency communication medium. Radio is the one means of communication that is considered to be ubiquitous and widely and freely used throughout the world. When the cell towers failed during recent hurricanes in the U.S., the radio signal was still there.

Radio is also a low cost solution for car makers, which may be why it isn’t taken more seriously. But car makers, like governments trying to send emergency communications, have a stake in the preservation of radio.

In every city around the world it is the radio broadcasters that deliver real-time traffic reports based on traffic reports from public authorities, cameras, helicopters and traffic spotters. Companies looking to deliver traffic, navigation and routing solutions ultimately turn to the broadcasters to source curated traffic incident information to support RDS-TMC and, for digital broadcast, T-PEG traffic services.

The challenge for broadcasters is to seize the reins of innovation in 2015. Consumers remain fans of broadcast radio content, but simply can’t find it as easily in connected cars as they once could.

The relevance of radio

The challenge for car makers is to recognize the importance of free over-the-air radio and preserve its privileged place in the dashboard. In others words, car makers should be careful not to become too dazzled by the app-oriented frenzy associated with connected cars. Sometimes, in fact most of the time, drivers just want to listen to the radio.

The fifth recommendation of the Digital Radio UK study was: “Develop an easy to use, open standard, hybrid (broadcast/IP) radio experience for new connected car dashboards.” This recommendation refers to efforts to “switch on” the FM chipsets already available on most mobile phones to enable such a hybrid experience. In fact, let's make sure those are digital radio chipsets. Companies such as Sprint and Emmis Communications are working hard to expand the adoption of hybrid radio in the U.S.

Through the adoption and promotion of digital and hybrid experiences, the humble radio can retake the innovation initiative in 2015. Radio needs to be everywhere that consumers are, including on their mobile phones. Car makers need to be reminded of the vital source of location information generally and traffic information in particular that the radio delivers – for free and without distraction. Without concerted effort we will continue to have the radio relevance discussions in 2015 that we had in 2014 – when we ought to be putting the debate to bed for good. The next headline should read: Radio Forever!


January 2, 2015 11:20 rlanctot

It seemed that Tesla could do no wrong in 2013 and 2014. Even when something went wrong, like its cars catching fire, Tesla turned it into an opportunity to demonstrate its technical advantages over the competition – with over the air software updates, of course. By the close of 2014, Tesla was running out of software tricks and was forced to impress with brute force hardware updates to vehicle performance.

But after doing its utmost to disrupt propulsion, safety, infotainment, manufacturing and distribution, Tesla will have to step aside for Uber in 2015. It is Uber that can brag of its $40B valuation now that it has set its sights beyond ride sharing to logistics and maybe more.

Uber is the envy of the technology community as it transforms transportation and explores the potential scope of its widening expertise. Uber is no longer solely interested in moving people. It is now selling and delivering things as well and moving into new forms of transportation (helicopters, boats) opening up intriguing possibilities.

Uber can tap into product delivery, navigation, routing and e-commerce. About the only opportunities not immediately on the table are mapping and search, but maybe Uber can have an impact in these applications as well.

While location has dominated discussions in the connected vehicle marketplace over the last few years, logistics represents the next level of added value. Uber’s knowledge of travel times between millions of destinations and its ability to gather and analyze probe data around the clock along with the ability to monitor the demand for transportation means Uber has the keys to unlock a deeper understanding of dynamic transportation requirements wherever it operates.

If Uber can add e-commerce to its menu of services it can offer a vertically integrated alternative to Amazon. In fact, as 2014 draws to a close it is odd to see Amazon so completely uninterested in transportation alternatives (aside from space travel) and the automotive market. Amazon doppelganger, Alibaba, in China, has acquired map maker Autonavi and entered the ride-sharing market with Kuaidi and talks about making its own cars. (Alibaba has even been enabling online auto sales and building OEM branded online shops.)

The leading browser in China, Baidu, has also entered both the map and ride sharing markets and has partnered with BMW to work on self-driving cars. Uber isn’t talking about building cars but the opportunity exists for Uber to use its app and probe data to build its own Waze-like map. An acquisition of TomTom by Uber could save the company a lot of trouble.

TomTom has been working with Apple for two years strongly suggesting that Apple has found the prospect of building its own map to be more challenging than it originally anticipated. Still, Apple has held off acquiring TomTom, leaving the way open for a competitor like Facebook, Alibaba, Baidu or, maybe Uber to buy TomTom and leverage its map for advertising, e-commerce, m-commerce and logistics opportunities.

HERE, the primary alternative to TomTom, is expected by many to ultimately fall into the hands of Microsoft, which has its own means and motives for scaling its Azure cloud services enhanced by HERE’s map data. But neither HERE nor Microsoft appear to be in any hurry and HERE, in the meantime, continues to add strategic deals with Microsoft competitors such as Samsung to extend its already strong automotive portfolio.

But Uber will be the fly in every player’s ointment in 2015. Uber’s secret sauce is its access to the growing millennial underclass rumored to be less devoted to owning cars than its forebears. For these users, Uber is the shizzle.

Every car maker will be transferring its awe and glorification of Google from 2-3 years ago to Uber in 2015, trying to tap into the next generation of automobile buyers. In anticipation, every Uber car will become something of a testbed not only for Uber’s dreams of product delivery and logistics, but also for a better understanding of consumer technology preferences.

Millennials are a tough crowd to figure out. They bring their own tech wherever they go, so it can be hard to impress them with yours. Uber users do need a credit card and smartphone to access the service, and those requirements fit nicely with the target audience of car makers. About the only value-adds an Uber driver can offer is the usual bottle of water and a smartphone charger.

The insights Uber possesses into the mass of Uber users and their rides, though, is invaluable and it spans most demographic segments. Uber can threaten Amazon’s domination of logistics and Google’s domination of location data interpretation – where users work, play, live etc. Like Tesla, Uber’s disruption is pervasive.

Working with Uber is an opportunity the auto industry cannot afford to pass up. This is why car makers have been setting up programs for Uber buyers to use Toyota, Audi or GM vehicles.

Uber is in position to be the analytics champ in 2015, if it isn’t already at the close of 2014. There’s a new king of disruption in the transportation biz. The only question that remains is whether Uber will be a responsible steward of customer data.


December 27, 2014 14:44 rlanctot

In California, Angelenos are complaining about “Wazers” – users of the Waze smartphone navigation app - taking local surface streets to avoid the presumably crowded freeways. Residents living near or between freeways are seeing an increased flow of meandering, side-street-clogging traffic attributable to Waze navigation.

In Brazil, female Paulistas eschew Waze because the app may steer them near or through a “communidade” or favela. Waze is incapable of protecting its users from potentially dangerous neighborhoods since it uses its own user-generated maps and does not discriminate safe from unsavory neighborhoods.

Still, people love Waze, even when it delivers some of the lousiest routing in the world.

Waze is incapable of predicting traffic conditions. In fact, the inability of Waze to accurately predict traffic guarantees its inability to deliver the best routes through and around traffic. And in my recent experience and the experiences of dozens of commenters in news reports, Waze almost invariably avoids main streets and freeways within dense metropolitan areas, such as Sao Paulo and Los Angeles – even when taking the freeway appears to be the better option.

But the Wazers aren’t complaining. The Wazers are in love with Waze and believe it can do no wrong… until it does.

Like my friend trying to visit me at the Omni Hotel in Los Angeles during the Los Angeles Auto Show who ended up in the Third Street Tunnel UNDER the Omni rather than in front of the hotel. My friend recognized that the app had screwed up because it had happened to him before while using Waze – telling him he had arrived at his destination when he was in the tunnel UNDER the hotel.

But Wazers keep coming back for more of this nonsense, even when Waze is taking them on a circuitous route through surface streets through obscure neighborhoods in major cities rather than on the highway. Wazers just assume that the Waze hive-mind is steering them clear of disastrous traffic conditions unfolding elsewhere.

But Waze has no predictive traffic function – only real-time data. So Waze can’t really provide the most qualified guidance for your route and appears to have little or no routing sense other than to avoid the traffic occurring in real time.

Avoiding freeways and main streets may be the best way to go, or it may not be. What is missing are algorithms designed to ferret out the best route taking into account historical trends, predictive models and major events in the area, weather, traffic lights, and stop signs and, of course, traffic incident data much of which comes from broadcasters such as iHeart Media Total Traffic and Weather Network or the Broadcast Traffic Consortium.

And discriminating between tunnels and surface streets might be useful too.

This is important to understand because routing is second only to traffic as the most important connected car (ie. telematics) application. Superior routing solutions will ultimately determine the winners of the navigation business and organizations such as HERE, TomTom and deCarta that understand this fact are working hard to distinguish their solutions.

But it is hard to explain the importance of routing and the physics and math behind predicting traffic and recommending the best routes. Even more difficult is performing this analysis in real-time, off-board, and delivering updates wirelessly to built-in navigation systems. It’s hard work and something that Waze isn’t capable of delivering.

Routing is a tough business and the best routing resides off-board – in the proverbial cloud. The best routing solutions require the best traffic data and the best traffic data has to be live and connected and combined with suitable maps. Waze rolls its own maps, which has its advantages until you are sitting in a tunnel underneath your destination.

Ask any fleet operator or EV driver about the importance of routing. They will tell you that the best routes are the either the fastest or the most fuel efficient - and you're going to want to know the difference. The best routes also integrate live traffic data AND predictive traffic information – these are requirements.

Waze offers nothing more than real-time probes, and some crowd-sourced inputs and is still working to enhance its integration of event data – a standard and fully-evolved element of competing solutions on smartphones, portable navigation devices and built-in systems in cars. We mindlessly follow Waze off of the highway and admire its cleverness in introducing us to the twisty shortcuts intended to shave minutes off of our routes, but time after time we see hints that these alternative routes are less than ideal.

Will we ever question Waze? Can we ever break free of its real-time spell and recognize the virtue of routing and navigation based on superior, high definition maps along with navigation derived from predictive analytics as enabled by TomTom, HERE and deCarta? Maybe. But until we do, we’ll keep zigzagging through obscure side-streets and looking for the Omni hotel in a tunnel.


December 25, 2014 14:09 rlanctot

“I find it difficult to believe that the seat belt can afford the driver any great amount of protection over and above that which is available to him through the medium of the safety-type steering wheel if he has his hands on the wheel and grips the rim sufficiently tight to take advantage of its energy absorption properties and also takes advantage of the shock-absorbing action which can be achieved by correct positioning of the feet and legs.”
- General Motors vehicle safety engineer Howard Gandelot, 1954

Howard Gandelot’s sentiments reflect a charming old common sense approach to vehicle safety. His perspective on driving safety was much like my own once was, forged before seatbelts and airbags were mandated. Safety was the driver’s and the passenger’s personal responsibility.

More than half a century ago my mother worked for Howard Gandelot, head of General Motors’ safety office in Detroit. It was after World War II and she only worked there briefly but it marked the high water mark of her corporate life before she embarked on marriage with my father, just returned from the Pacific theater, and raising a family.

My mother’s family came north to Detroit earlier in the previous century to find jobs in the auto industry. My father’s father was a factory worker on the assembly line, possibly at one of Ford’s facilities. I’m not sure.

All of this came to mind while visiting my 88-year-old mother this week. I knew she had worked at GM briefly in the 1940’s, but I hadn’t connected the fact that she worked in the safety office.

We generally don’t think about car companies having safety offices in the first half of the 20th century. But they did. In fact, the Gandelot quote (above) came eight years after my mother left GM.

Auto safety is on my mind, due to the latest headlines:

- Highway fatalities decline - 2013 Motor Vehicle Crashes: Overview (DOT HS 812 101), After an increase in motor vehicle crash fatalities in 2012, fatalities on U.S. roadways in 2013 resumed the decline that had started seven years prior. Despite the decline in fatalities, 32,719 people died in crashes on roadways during 2013, down from 33,782 in 2012. The number of people injured decreased in 2013 as well, falling from 2.4 to 2.3 million. – NHTSA.gov - http://tinyurl.com/mpjykzt
- U.S. recalls top 60M for first time - http://tinyurl.com/k9mjqgm - Bloomberg
- “Growing list of safe cars belies fears in record recalls” – “The insurance industry’s list of cars and trucks that do the best job of keeping owners alive in a crash jumped 82 percent this year against a backdrop of U.S. recalls of older models that were killing their passengers.” – Autonews.com
Automotive safety permeates all of our lives. As I drove my mother to a doctor’s appointment this week she offered me a steady stream of safe driving tips. Well she might, since her driving record is better than mine!

Howard Gandelot is an interesting character to ponder in this context. He was a safety engineer before these positions were fully exposed to class action lawsuits and significant government oversight. His common sense perspective on vehicle safety contrasts with today’s sensor- and algorithm-based analytics and big data approach.

We have come a long way from Howard Gandelot’s time, or have we?

Highway fatalities fell to the second lowest level ever last year and vehicle recalls – at more than 60M, according to a Bloomberg investigation – doubled the previous record. What are we to make of this good news, bad news proposition. The recalls were driven in large part by the GM ignition switch failure (now believed to be implicated in 42 deaths, and counting) and the Takata airbag recall – impacting nearly every car maker..

As sophisticated as we think we have become in enhancing vehicle safety, the Takata recall shows that we are still operating in the realm of trying to save lives mechanically – ie. with an airbag deployment – rather than intelligently with sophisticated analytics. Metaphorically, we are like the apes throwing bones at the monolith at the start of “2001: A Space Odyssey.” We are still only working with primitive tools.

We have the know-how and the data to rapidly advance our understanding of the causes of airbag deploying crashes, but the data, the tools, those bones being thrown by the apes, are not being put into the proper hands for analytical purposes. A colleague of mine, Chris Schreiner, co-authored a study while at GM, that correlated airbag deployments with the use of OnStar personal minutes.

The study found relatively fewer airbag deployments when OnStar users were using personal minutes than when they were driving without using personal minutes. The point is, this is precisely the kind of research the industry and government should be collaborating on.

Honda was penalized earlier this year for insufficient reporting of airbag deployment crashes. It seems odd that car companies are responsible for reporting these crashes when public authorities such as police and emergency medical responders are normally involved and capable of collecting relevant data.

Auto collisions are traumatic events and ought to be treated as such. They occur on public roadways and, as public events, the observed outcome ought to be collected – including, possibly, on-board or even off-board data relevant to the crash.

Industry critics frequently complain that it is impossible to obtain real-world crashworthiness data on cars – but this information is available if we choose to collect, analyze and disseminate it. The past year saw NHTSA’s safercar.gov Website gain wider attention for its collection of individual vehicle recall and complaint information – but more must be done.

The Insurance Institute for Highway Safety continues to send mixed messages about vehicle safety. Following the report of a record year for recalls of cars to fix life-threatening vehicle problems, IIHS reports a substantial increase in the number of cars found to be safe for protecting vehicle occupants. But what if those systems fail?

IIHS has taken the place of Howard Gandelot in fatuously reassuring the general public that cars are safe, all is well and we have nothing to worry about. IIHS’s work should not be mistaken for consumer advocacy.

Annual highway fatalities may have declined in 2013, but the U.S. still slaughters enough people on the roads every year to rank the country as #4 in the world – with all of the societal and personal cost that that implies.

Since Howard Gandelot’s time we have discovered the life-saving qualities of the airbag. But with the Takata recall even airbags are failing us.

A more reasonable position will be to put big data to work immediately in the service of saving lives. Now is the time to start gathering accident data directly – rather than indirectly through auto makers – to identify potential flaws and failures in real-time.

- We have the technology to identify risky driving behavior both in real-time and retroactively.
- We have the ability to alert drivers entering dangerous turns (in real time) that they are on a fatal trajectory.
- We have the technical ability to alert drivers approaching intersections that another driver is approaching the intersection in a manner that will cause him or her to violate the red light.

We have the ability, retroactively, to understand what caused an accident that triggered an airbag deployment. Or, maybe, we can use the available data to determine that the airbag should not have deployed at all or should have and failed to.

The bottom line: We have left the Howard Gandelot age of “common sense” safety and entered the age of “big data” safety. It is time to take off the blind fold, stop throwing the bones at the monolith and embrace the opportunity that big data analytics presents to prevent accidents from occurring or recurring. This is the new common sense.


December 21, 2014 09:57 rlanctot

One of the biggest highlights of the upcoming CES show in Las Vegas, put on by the Consumer Electronics Association the first week in January, will be the focus on the Internet of Things. While most journalists and attendees will focus on the morphing of televisions into home hubs, they might miss the emergence of IoT as the replacement for vehicle-to-vehicle communications.

Astute industry watchers will have picked up on the Volvo bike helmet story on Engadget (http://tinyurl.com/q5kjwkc - ‘Volvo's bike helmet concept alerts riders and drivers to each other’). But even these devotees may have missed the implications of a bike helmet communicating with a car through an app.

The significance, quite simply and clearly, is that communications with vehicles will not require an entirely new wi-fi-like network as envisioned by the U.S. Department of Transportation and foreign and domestic ITS executives. Communication with vehicles and between vehicles and other vehicles or infrastructure is being made possible via either connected smartphones or embedded telecom modules, a la GM’s OnStar service today.

In fact, car makers are increasingly turning to existing wireless networks to enable this kind of communication not only to identify and avoid hitting pedestrians, as in the Volvo use case, but to identify and avoid hitting other cars or to be alerted to road hazards. In fact, the concept of driver alerts is ideally suited to the kind of use cases envisioned for the Internet of Things.

This communication between cars and smartphones and connected “things” has a precursor in Waze. Waze was first to popularize and mass market the concept of crowd-sourced traffic and map data along with the concept of driver alerts regarding everything from potholes, to speed traps to traffic.

The ITS V2V crowd has co-opted the Waze concept with their own idea of “swarm intelligence” – a phrase first used in the late ‘80’s – right around the time the V2V folks began latching onto 802.11p Wi-Fi technology to create a proprietary, industry-specific means of communication between vehicles. By now, of course, this concept has been superseded by the proliferation of powerful smartphones, cloud connectivity and analytics, and cheap camera and wireless sensors.

So, while consumers are just becoming aware of IoT and its application and monetization possibilities (Harbor Research: $1T/28B things; IDC: $8.9T/212B things; Cisco CEO: $19T/everything), the ITS crowd has kept its head down insisting that V2X communications will require new hardware, new infrastructure and protected spectrum.

Forget that.

CES 2015 signals the ultimate demise of existing V2X approaches and the sooner this is seen the better for the safety of drivers and other users of roadways.

V2X via dedicated spectrum is dead man walking thanks to IoT.

Different forms of connectivity are better suited to different applications. In the Volvo use case the helmet-wearing bike rider must be using an application on his or her smartphone that connects to the “cloud” which then communicates to a Volvo vehicle via the embedded modem.

But what if the car is not a Volvo and what if the car does not have a built in modem. In this case, the clear path to directly communicate an alert to the driver is via that driver’s connected smartphone.

But the smartphone alert only works if the phone is connected in the car. While this may require some behavioral change to convince consumers to connect their phones in cars – PHYSICALLY via a cable – it also means that the connected phone will communicate through and with the vehicle.

This is not to be confused with Google’s Android Auto smartphone connection or Apple’s CarPlay. Only proprietary connectivity systems like Ford SYNC or GM’s MyLink will enable a deep integration with the vehicle systems for driver alerts. It is worth noting that neither Google nor Apple have proposed this kind of safety alerting for their smartphone integration systems – just another reflection of their disconnection with automotive industry priorities.

The Volvo smartphone-to-car connection is using Ericsson’s cloud technology. But what if the helmet could communicate directly with another smartphone or car?

The vision of IoT is to enable connectivity and the resulting analytical intelligence to occur directly between devices or via devices and infrastructure to other devices or to the cloud and, thereby, to other devices. The related analytics can also occur on the device, in the cloud or at the edge of the network – in the infrastructure.

In fact, emerging wireless technology, LTE Advanced and (in 3-4 years) 5G, will enable direct device-to-device communications meaning the bicyclist’s helmet could communicate directly to a smartphone connected in a car (or the embedded modem) to alert the driver to the bike rider’s proximity. But Ericsson and Volvo are delivering this kind of communication today, at least for appropriately equipped Volvo vehicles and Daimler, too, is using connected smartphones and embedded connectivity to enable inter-vehicle communications of highway and traffic conditions.

BMW demonstrated infrastructure-to-vehicle communications using cellular technology at a recent conference in Detroit. And even GM and Honda have looked at smartphone-based V2V communications.

IoT technology will open up communication between cars and everything else: smartphones, wearables, homes/buildings, and infrastructure. It is this vision which is rapidly rendering moot the US DOT’s plans for a purpose-built, dedicated wi-fi network. It is increasingly becoming clear that we don’t need this network. It will take too long and cost too much to build, even if the capacity, interference, cost and security concerns could be resolved.

It appears that the auto industry is waking up to this opportunity. A November 11th connected car working session of the Linux Foundation’s AllSeen Alliance included participants such as Daimler, BMW, Volkswagen/Audi Research, Continental, Panasonic Automotive, Nissan, General Motors, and Sony.

To catch up with the latest developments, prime CES exhibitors (in January in Las Vegas) include Muzzley at the Sands Hotel and Qualcomm in the Central Hall. The list of AllSeen Alliance members at CES includes a handful of automotive players, but the complete list of IoT exhibitors appears here: https://allseenalliance.org/event/ces-2015.

Meanwhile, executives at Technicolor are seeking to create a connected car project and working group within AllSeen Alliance after having coordinated the recent working session. Bottom line: V2X is here today manifest as IoT and will be visible at CES.