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

August 19, 2014 15:24 rlanctot

Yesterday I said buh-bye to UBI car insurance. UBI is the acronym given by the insurance industry to the service based on a tracking device being installed in the insured driver’s car. It stands for usage-based insurance.

I had been a State Farm customer for approximately 15 years. And my participation in State Farm’s UBI program ought to have cemented our relationship. Instead, the UBI program was irrelevant.

This powerful new customer connection – UBI - did not lead to a change in how State Farm interacted with me. And that failure is just one of many reasons UBI insurance is struggling to achieve wider traction.

Progressive was the pioneer of UBI insurance and its Snapshot modules have been installed (temporarily) in more than a million vehicles for the purposes of determining an optimized insurance rate based on driving behavior. The program was conceived to:

1) Provide a more accurate calculation of risk

2) Provide a conditional manner for insuring high-risk drivers

3) Attract and poach low-risk drivers from competitors

4) Reduce customer churn

5) Lower the costs associated with claims management*

6) Speed up claims*

*These aims are not served by Progressive's Snapshot

In the U.S., as in my case, the customer-facing purpose of UBI insurance is to get a deeper insurance discount. Most programs – currently offered by dozens of insurance companies around the U.S. – offer a 5% upfront signing discount with potential discounts of as much as 30% on premiums. Those discounts are based on driving behavior scores which are derived from devices affixed to the driver’s vehicle. There are also smartphone-based solutions.

As I discovered, bigger discounts are available simply by changing insurance carrier. Conventional insurance programs with features such as Liberty Mutual’s “accident forgiveness” and Nationwide’s “vanishing deductible” resonate more powerfully, especially since they tend to be uniformly offered throughout the U.S.

I participated in State Farm’s Drive Safe and Save program for the past year. I have written and spoken about my experience frequently. I left the State Farm program because Liberty Mutual – approaching me as a result of my ownership of a BMW – offered to insure my car and my wife’s car at approximately 50% of the rate I was paying State Farm.

This is no reflection on the UBI program offered by State Farm. The UBI device from State Farm (which I will now have to mail back) worked just fine, the consumer portal worked just fine, the annual discount seemed generous ($200+), and the added value services worked as promised. And I had no complaints about my State Farm agent, with whom I now have a strong relationship.

In fact, I would have signed up for Liberty Mutual’s Right Track UBI program if it were available in Virginia, which it is not. It is available in 22 other states.

In the end, the conventional policy from Liberty Mutual represented too good of an offer to pass up. The sticking point, in retrospect, appears to have been an accident/claim from three years ago, the details of which beyond the amount (~$2,000) I can no longer remember. That accident is still within the three-year window that is keeping my State Farm rate in an unacceptable range.

But the bigger issue is the lack of transparency in the insurance industry. And that lack of transparency extends not only to the driver/customer, but also to the broker. This is where the industry needs to be rewired.

My State Farm agent noted that most of what goes on in State Farm’s underwriting black box is completely opaque to him. He is aware that factors affecting rates are changing on a daily basis, affecting the quotes he is giving, but he does not understand why and he has no visibility to when.

In his immortal words: “Everything is being thrown at me except what I need.”

So State Farm failed me and is failing its own agents.

State Farm had a privileged relationship with me that is now lost. The company had a module in my car that was reporting on my driving behavior – speed, time of day, acceleration, turns, and amount – whenever I or my wife drove our vehicles.

The information was shared with me on a portal with some limited interactive capabilities. State Farm also had a mobile app, though I never successfully used this.

The presence of the Drive Safe & Save device in my car opened a door for State Farm to fundamentally alter the consumer-insurer relationship. State Farm could have:

1) Advised me as to how to improve my driving and thereby increase my discount.

2) Compared my driving to the aggregated driving behavior of other drivers.

3) Offered me rewards for better driving.

4) Notified me of negative factors – tickets, accidents – impacting my rate.

Needless to say, State Farm did not do any of this. It was a missed opportunity. But that failure was felt by my agent as well.

I would have benefited from a graphic on my bill or Web portal indicating when particular accidents or infractions might expire from my underwriting score. My agent, too, could benefit from notifications regarding scoring updates impacting his customers.

A driver’s given rate might change on a daily basis, according to my agent, though he has no way of easily seeing that. Can you imagine your auto insurance agent calling you to let you know your rate had gone down? That is a relationship changer.

(The Liberty Mutual agent was able to call up the Clue Report to access my claims history. The State Farm agent could only refer to my file – he did not have access to the same resources as the Liberty Mutual Agent.)

Consumers are entitled to see what factors are affecting their rate. Driving history, credit record, and location are the main factors, but there are others. Just as consumers can access their credit reports, their driving history should be integrated into their insurance portals and bills. I felt a little strange that I had to ask my broker about my own driving history. If this information is publicly available insurance companies should help to make it more accessible to consumers.

Full disclosure of the factors, if not the balancing of those factors, should be required and might contribute to creating a higher degree of customer trust. But the industry won’t be able to fix this trust factor until agents are treated better.

Car insurance providers such as State Farm and Allstate employ upwards of 15,000 independent agents interacting with customers on a daily basis. These agents should have customer dashboards to alert them to driving-related events that will impact their customers’ rates, and they should also have live feeds to alert them to changes in rating criteria that impact customers’ current rates.

For the most part, insurance agents are flying blind. This blindness contributes to the lack of consumer trust and undermines the ability of the industry to adopt new technology, such as telematics, for tracking customers and delivering a service based on a more accurately calculated risk assessment.

And make no mistake, the adoption of telematics will require trust.

The game is changing to one of connected drivers. It isn’t enough to have customers plug in a box and get a 5% discount – especially if they can save more by switching providers. A new connection calls for a new paradigm of customer engagement. State Farm missed the connection.

I agreed to take my State Farm agent’s call in December when my three-year-old accident will no longer mar my driving record and my discount. We’ll see if State Farm can re-connect.

Please join me @ Insurance Telematics Sept. 3-4 at the Radisson Blu in Chicago -

August 12, 2014 16:12 rlanctot

This is a follow-up to the original blog:

So let me make sure I get this right. Hackers Charlie Miller, security engineer for Twitter, and Chris Valasek, (left) director of security intelligence at IOActive:

1) “Hacked” into a Toyota Prius and a Ford Escape.

2) Shared the details of their “exploit.”

3) Published a list of the Top 20 Hackable Cars.

4) Introduced a device to detect and prevent hacks.

Now let’s break it down further. Miller and Valasek’s work:

1) Embarrassed Toyota and Ford and, by extension, the makers of the cars on the Top 20 list that received failing grades.

2) Used their non-automotive expertise to broadly cast doubt on all vehicle security.

3) Scared and misled consumers about the current state of vehicle security and the implications for drivers and vehicle owners.

Miller and Valasek shared nothing about the automotive security work of groups such as:

1) Standards-setting bodies - ISO, SAE, JEITA, ITS Forum, GenIVI or AGL

2) EVITA – E-Safety Vehicle Intrusion Protected Applications

3) PRESERVE –Preparing Secure e-Vehicle-X Communication Systems

4) DARPA’s HACMS – High-Assurance Cyber-Military Systems

5) Alliance of Automobile Manufacturers/GlobalAutomakers – plan to create an industry sector information sharing and analysis center.

6) Battelle’s CyberAuto Challenge

7) DEFCON – (Automotive industry participation)

8) escar – Embedded Security in Car conference.

This is to say nothing of the ongoing work taking place across the entire automotive eco-system.

Everyday, hardware, software, semiconductor and service provider organizations are seeking and implementing solutions to automotive security challenges from proliferating wireless interfaces down to board-level silicon.

As alcoholics say, if you are trying to solve the problem, you have to start going to meetings. There are plenty of organizations with plenty of meetings, gentlemen. Pick one.

Lobbing hand-grenades over the wall to prove the existence of security weaknesses is unhelpful, self-serving and pointless. Confusing and misleading consumers about the true scope of the problem and potential remedies is reprehensible.

August 10, 2014 16:31 rlanctot

J.D. Power waded into the swamp of automotive voice recognition technology last week at the Center for Automotive Research’s Management Briefing Seminar in Traverse City.  JDP’s executive director of driver interaction, Kristin Kolodge, presented slides and videos to show JDP’s assessment of the abysmal state of VR today.

Kristin said JDP’s annual Initial Quality Study of vehicles sold in the U.S. revealed VR tech as the most common type of “malfunction.”  VR was to blame for one-third of infotainment system failures – which is significant since infotainment systems have emerged in the past few years as the single biggest source of failures in new cars.

Kristin’s prescription, according to the report on her talk in Automotive News, was to get back to basics – that automakers should give up trying to add new features.  This recommendation alarmed me because the industry is actually on the verge of a major industrywide upgrade to natural language speech recognition and this is no time to turn back.

Recognizing that it had a problem on its hands with VR technology, most car makers have been turning in the direction of the skid.  Car makers see that they need to do better and that doing better means bringing automotive grade speech recognition systems to cars that adapt to humans rather than forcing humans to adapt to them.

There are several problems with VR today and they include:

Overly specific menus and poorly conceived architectures

Attempts to use voice recognition where it is an inappropriate interface

Multiple on-board speech recognition systems

Voice interfaces that work with some apps and not others

Confusing cues

Speaking to an inanimate object is an unnatural act, so there is no surprise that getting consumers to change their behavior is a big step.  Ford took the biggest step by making speech recognition the focal point of the original SYNC system.  But Ford changed its VR architecture and expanded the vocabulary with SYNC Gen 2 with disastrous results.

The problem is that once consumers have had a bad experience, it is tough to win them back.  And winning consumers back to automotive VR is important because VR is a powerful tool for combatting driver distraction.

Unfortunately, VR systems on the road today - most of which were designed or created three yeras ago - actually create distraction.  So let’s quickly review where we are headed with speech recognition:

NLU is the future of VR in the car

Whether you look to AT&T’s Watson or Nuance’s Dragon Drive or to VoiceBox’s conversational recognizer, VR tech is rapidly becoming a more natural experience in the car thanks to natural language understanding (NLU).  It is true that Apple’s Siri and Google Voice work impressively on mobile devices held close to the mouth, but in the car, automotive grade systems optimized for the automotive environment and automotive use cases are best.

Learning and Personalization

VR suppliers are increasingly integrating abilities into the NLU systems that introduce learning capabilities.  A standout in this area is MeMeMe Mobile which personalizes speech recognition to the speaker for use within and beyond the car.

Application Focus

Drivers want a reliable VR solution for hands-free access to telephony, navigation (destination entry!) and audio.  All other functions in the car – such as HVAC – are best handled with other types of controls.  But appropriate integration of VR technology is key.  You may use voice to look up an audio track, but a manual control to increase the volume.

Reducing Distraction

Luxury cars are already reading text messages and emails and allowing drivers to respond in a hands-free or, preferred, an automated manner.  The influential California legislature has considered banning this type of functionality, but, if implemented properly and without legislative interference, such a capability could be a useful distraction mitigation tool.

JDP’s IQS study is an important bellwether for the automotive industry.  But let’s not forget that the perennially grumpy users responding to JDP’s study are using three-year-old technology.  Advances in automotive VR are closer than they may appear to be when looking at today’s new cars.

August 9, 2014 14:15 rlanctot

The Annual Traverse City Management Briefing Seminar quietly concluded this week without shedding any newsworthy light on the future of alternative powertrains, the realistic prospects for autonomous vehicles or the ongoing impact of China on vehicle design and production.  Also missing was a presentation explaining how the auto industry was going to overcome the worst year on record for vehicle recalls.  And no one mentioned the ongoing carnage on U.S. highways – nearly 100 daily fatalities.

If it weren’t for a stirring, from the heart and straight from the shoulder, speech from Fiat’s Sergio Marchione (who was coincidentally briefing financial analysts away from the event), I am sure the attendees will have spent much of their time sleeping off their cocktails or working on their golf strokes.

Full Automotive News coverage can be found here:

I got to thinking about Traverse City because I am preparing to moderate a panel at an overseas event and the sponsor and operator of the event wants to put its customers on the panel I am moderating.  My thought was that it will be more interesting to put the customers’ customers on the panel – they are more likely to speak freely.

Free speech was in short supply at Traverse City, at least until Marchione stepped to the lectern.  Marchione spoke freely about wanting to claw back the profits his suppliers are bragging about every quarter in their financial reports.  This perspective was in conflict with the more palliative comments of those same suppliers, many of whom spoke at the event of a new age in the automotive industry when auto makers will no longer hammer their suppliers chipping away at those profit margins.

These auto industry suppliers – nVidia and Delphi notably among them - were surfacing an issue which has crept to the forefront of the automotive industry and which the industry itself is still struggling to manage.  The average car now has more than 100M lines of code on board.  The software content is gradually coming to eclipse the hardware content in the car. 

Sourcing software is a far more complex exercise than sourcing hardware.  What are you buying as a car maker?  Bits and bytes?  Software code?  Do you own that code or do you rent it?  Does your supplier own that code or is it open source?  How do you maintain and/or replace that code?  And where does liability reside?


For years the automotive industry has been pursuing a strategy of segregating hardware from software, while suppliers have been fighting to keep hardware and software firmly stitched together in competitive bids. 

By selling hardware and software as a package, Tier 1 suppliers may feel they have more pricing flexibility.  Now, even Harman has acknowledged the ascendance of software.  Harman CEO Dinesh Paliwal recently told Bloomberg in an interview that his intention is to reduce the companies hardware manufacturing activities and shift its focus to software, which already represents 75% of Harman's revenue. ttp:// - Harman Stakes Future in Software as Autos Become Smarter)


Unfortunately, car makers are increasingly determined to segregate these two pieces.  The net result is that contract manufacturers have emerged to compete with and sometimes partner with Tier 1s while software only integrators have proliferated.  Tier 1s that might have played fast and loose in their bids – offering “integration” for “free” may have painted themselves into a corner. 

All suppliers are putting a price tag on integration these days.  In fact, integration has become a “thing.”

Tier 1 suppliers from Bosch and Continental to Delphi, Magna and Denso are caught up in this battle to either preserve their hardware/software packages or create separate software teams to pursue RFQs with segregated hardware and software requirements.  But it is more complex than that, as auto makers have sought to take ownership of the software code and related intellectual property.

Missing from the Traverse City event was a voice from the software side of the industry.  Stuffy old Traverse City

(50 years running) missed the email once again – nary a word was heard of open source software, Google/Android or Linux.  The most likely source of such a perspective was John Ellis of Ford, who participated on the “Designing for Technology and the Customer” panel.

John is a truth talker, but I’m not sure the Traverse City crowd was ready for two double shots of espresso in one week.  Marchione surely left a few headaches and upset stomachs in his wake.

Just as Google was not represented at the Traverse City event there was also no speaker from a wireless carrier, a Chinese car maker, Tesla Motors, a leading dealer organization (with the possible exception of Joe Carlier, senior vice president of Penske Logistics) or a single car owner – or even a victim of a car crash resulting from a vehicle defect.

The rising influence of software is transforming the automotive industry.  To preserve its relevance, the Traverse City Management Briefing Seminars needs to integrate a powerful voice for the role of software in vehicle design and operation.  How about a panel on the role of GenIVI, AUTOSAR and model-based software?  How about a panel on the challenge in finding enough programmers to fulfill the industry's requirements.  Even powertrains come with engine controls, after all.


It would also be helpful to include a voice from the wireless industry – since vehicle connectivity and software updates are rapidly becoming de rigueur.  Tesla anyone?

We’ll see if the Center for Automotive Research can reverse this Traverse City travesty.  If it continues on its current trajectory, your time on the Upper Peninsula, if you go next year, will be better spent on the golf course.

August 6, 2014 13:11 rlanctot

Last December two hackers, Charlie Miller, security engineer for Twitter, and Chris Valasek, (left, bottom) director of security intelligence at IOActive, tried their hand at standup comedy with the unlikely topic of automotive security. Like Comedy Central’s Key & Peele (left, top) they sought to mine the process of vehicle hacking for yuks and I can honestly say, they were pretty successful. The proof:

This week the Miller and Valasek comedy team have released their roster of “the world’s most hackable cars” as reported in InformationWeek: And they will present today at Blackhat 2014 in Las Vegas (“A Survey of Remote Automotive Attack Surfaces - where, at the conclusion of their talk, they are expected to demonstrate a device, created from $150 in parts, to detect and prevent hacking in a car. The device plugs into the OBDII port and monitors vehicle network traffic.

There are two kinds of security presentations: The ones that scare you half to death and send you running to cancel your credit cards, and the ones that clearly delineate the extent of the problem and the known solutions.

But Miller and Valasek have taken a third path morphing from comedians and fear-mongers into pitchmen. Key & Peele cum Ron Popeil. Their vision of vehicle security apparently boils down to an aftermarket device. No marketing or sales plans have been announced.

There are a few things wrong with the Miller and Valasek message:

#1 They attempted to show how easy it was to “hack” into a vehicle network to access vehicle controls such as brakes and steering. For the purpose, they chose a Toyota Prius and a Ford Escape equipped with parking assist technologies.

The reality is that their “hack” – much of which could have been achieved with off-the-shelf diagnostic tools – required months to achieve as was made clear in the video. In retrospect this is oddly reassuring rather than terrifying. There was nothing simple nor was there anything remote about their vehicle “security breach.”

#2 Their roster of most hackable cars appears to be based entirely on whether or not the car has integrated controllers for safety and connectivity. In their estimation, the more segregated the vehicle systems are from each other and from connectivity, the better.

The reality is that vehicle systems are becoming increasingly integrated for the purpose of enabling autonomous driving and other safety-related applications including diagnostics. Commensurate with this integration has been a much greater focus from car makers on the security of on-board systems.

#3 They offer a device for monitoring for vehicle intrusions. But vehicle security is not an aftermarket product. If anything, the attachment of an aftermarket device is more likely to increase rather than decrease system vulnerability.

Vehicle security is a multi-layered proposition encompassing everything from the semiconductors to the wireless connections to the on-board networks all the way down to individual ECUs. Vehicle security is a philosophy that takes into account hardware and software and even leverages wireless connectivity for authentication and access (soon) to public key infrastructure and software updates.

The industry is rapidly moving toward a more robust gateway-ed and firewalled approach to security, but one that will enable off-board to on-board communications. Is remote control possible – sure – but it is not easily achieved, as Miller and Valasek have shown - chafed knuckles and all.

Miller and Valasek have helped to raise awareness of the security problem. Where they have failed is raising the understanding of the solutions to the problem which exist and are being implemented from suppliers as varied as Harman International and Covisint to Intel, NXP, Freescale, AMV Networks, QNX, and Cisco. Maybe, like Key & Peele, they need a little dose of Liam Neeson.

August 4, 2014 16:26 rlanctot

FM radio is locked in the car. And the wireless industry holds the key.

The once-ubiquitous FM radio is today primarily enjoyed in the car, where a growing variety of audio consumption occurs. But stored and streamed content on mobile devices is steadily marginalizing the broadcast portion of in-car listening, sucking the audience lifeblood out of the medium.

This phenomenon wouldn’t be such a problem if mobile devices could access broadcast radio in the U.S. The sad reality is that the semiconductor chips required to receive broadcast signals are already built into most smartphones in use in the U.S., but the wireless carriers – with one exception – have chosen not to enable access to broadcast radio. (Outside the U.S., broadcast signals are routinely accessed via smartphones.)

Sprint is the one exception in the U.S., offering NextRadio (an app created by Emmis Communications) to let its smartphone users (for specific models) receive broadcast signals by using the headphones as an antenna. NextRadio provides an app for free over-the-air listening to local radio stations – not to be confused with apps such as iHeartRadio or TuneIn Radio which enable listening to in and out of market radio stations via an IP stream.

The NextRadio app cleverly sorts the local stations by genre, tracks listening history and allow functions such as “like,” “dislike,” “share,” “buy,” and “save.” Switching between local stations is also simple. HD stations are not available.

Coleman Insights recently released the results of a consumer study intended to test the messaging and positioning of NextRadio with a representative sample of potential U.S. users. Results of the study are here: RAIN Newsletter analysis of the results are here:

What is clear from the study is that the participants in the study were impressed with the app and seemed to find the interactivity appealing. Participants in the study were shown the following video introduction: and then asked for their reaction.

The Coleman study focused on users under 40. It is apparent that those of us (wink wink) over 40 are blind to the fact that radio is no longer ubiquitous and is, for some, hard to find outside the car.

Coleman’s conclusion from the study is that NextRadio be positioned as “FM radio on your smartphone.” A secondary message or benefit is its use of the broadcast signal so that it has much less impact on battery use and only a tiny impact on data usage.

It so happens that one of my sons just picked up a Sprint phone and was able to demonstrate the app to me. Like the Coleman survey respondents, he found the app to be compelling and clever, but he was even more interested in the six-month free trial of Spotify that Sprint is offering and the $5/month friends and family plan available after the commercial-free trial period. Ergo, even Sprint is distracting potential NextRadio users from the free FM tuner available on some of its phones.

The Coleman study found a variety of complaints regarding existing streaming audio alternatives including stale play lists, limited opportunities to discover new content and too many ads. What the Coleman study failed to take into account is the range of listening options that are continuing to emerge including Amazon Prime, Google Play, Aupeo!, Apple’s iTunes Radio, Pandora, Spotify, Beats, Rhapsody, SiriusXM, YouTube, TuneIn Radio, iHeartRadio, Rdio, Slacker and Aha Radio.

From the study, it was not clear how much audio usage the participants engaged in or across what range of apps, devices or scenarios. Audio consumption occurs at home, in the car, at work and on the go. Apps and services are accessed in the car, on mobile devices and on desktop and portable computers.

User preferences are determined by cost, presence or absence of commercials, on-demand content vs. curated or automated stations, playlists, social network integration and availability of new music, among other factors. Most of these elements were ignored in the Coleman study in order to focus solely on the concept of FM radio on a mobile phone.

“FM radio on your smartphone” is a reasonable marketing statement around which broadcasters can rally, but the underlying message for me is “FM radio for everywhere else.” NextRadio is facing multiple hurdles including recruiting wider wireless carrier support and the long-shot proposition of bringing Apple on board.

Of course, if Google and Apple are really determined to displace the in-vehicle car stereo they should be taking a closer look at NextRadio. For Apple it is a simple enough decision to implement. (Apple phones have FM demodulator capability today but lack filters and antennas to enable FM to work.)

For Google, it amounts to nothing more than adding NextRadio tech to AndroidOne reference designs. At the recent Google I/O event, indications were that AndroidOne will have the necessary FM radio silicon, but the full scope of enabling technology to be included was unclear.

RadioDNS in the UK is working on setting standards for enabling a full hybrid (IP + broadcast) experience on smartphones and in cars. Auto makers are working with RadioDNS and ETSI to make this happen and significant progress has been made.

Starting with Nokia, handset makers and carriers outside the U.S. long ago decided to enable FM reception on handsets as a tick-box feature. The next step is to make the broadcast signal available on handsets as a resource either for standalone apps or as an accessible resource for other apps.

It is possible that Nokia's historical weakness in the U.S. contributed to the lack of interest in radio integration on mobile phones. RadioDNS is working closely with Emmis to merge the NextRadio and hybrid radio experiences. Today, NextRadio advocates must customize the app to work on a phone-by-phone basis.

In the meantime, NextRadio is stuck in the awkward position of trying on its own to liberate the radio from the car – the one venue where radio remains simultaneously strong and imminently at risk. And to escape the car, NextRadio is dependent on the support of the wireless industry – which is way more interested in enabling streaming audio services than facilitating free over-the-air reception of content.

NextRadio has the power to turn radio into a much more interactive medium. But the wireless industry holds the key to unlocking that opportunity. The broadcast industry exchanged a substantial amount of advertising airtime to bring Sprint on-board. Is there enough inventory to bring the rest of the wireless industry along?

July 31, 2014 22:04 rlanctot

Given the relatively high cost of electric vehicles relative to conventional internal combustion cars, car makers are constantly looking for ways to enhance the attractiveness of this privileged purchase. BMW is the first car maker to take the novel approach of creating a usage-based insurance product tied to limited annual mileage to reduce the cost of insuring a BMW i3 EV.

The strategy has not seen widespread adoption and only The Hartford in the U.S. is offering an EV insurance discount – 5%. The rational for offering such a discount seems sound given the limited range of most EVs and, therefore, the expectation of less frequent use and lower mileage.

Perhaps the most unusual aspect of the BMW UBI offer in the UK for the i3 is BMW’s inclination in the U.S., thus far, to avoid usage-based insurance programs as antithetical to the much touted “Ultimate Driving Experience” that a BMW offers The thought process is that rapid acceleration and deceleration and driving at the edge of the vehicle’s performance range should not produce an insurance penalty.

By limiting the UK program to mileage only, though, both BMW and its insurance partner have avoided an onerous intrusion into the customer’s enjoyment of the car. In fact, the mileage-based program is more or less the same as mileage-based insurance discounts from GM and Ford in the U.S.

Since late last year BMW has been offering a usage-based insurance product, called FlexiMile, to purchasers of the BMW i3 in the UK in cooperation with Allianz, the insurance carrier underwriting the offer. On its customer Website BMW states: “In addition to our standard comprehensive BMW Car Insurance, BMW FlexiMile insurance has been developed exclusively for BMW i vehicles. This new insurance product lets you enjoy a lower premium (approximately 25% discount) based on an annual mileage of 5,000 miles.

“There is no requirement to do anything – using BMW ConnectedDrive technology the BMW i3 will intelligently track the distance travelled, producing a monthly statement to keep you up to date on your mileage. Any mileage over 5,000 miles will be charged on a fixed rate-per-mile basis, and you’ll be able to switch to BMW Car Insurance – Unlimited miles, should it be more cost effective for you.”

Strategy Analytics reached out to BMW for some clarification on the program and the assumptions behind it and its objectives. Here are the responses from BMW:

  • Why offer this only for BMW i vehicles?

The technology to capture and send customer mileage via ConnectedDrive is standard in all customer supplied i3’s.  This technology is not available as standard for BMW or MINI combustion engines at this time (in the UK), which is why we had the opportunity to develop FlexiMile for i3 in the first instance.

The i3 has been a new and exciting customer offering - FlexiMile was developed (amongst a number of new customer offerings and packages) to supplement this advancement in the BMW range.

  • Isn’t 5000 miles pretty low, even for an EV? That’s only 100 miles a week.

FlexiMile was named as such so that it is flexible to customers’ needs.  Research carried out in the development stages of this product identified that 5,000 miles was a suitable threshold for the majority of customers using an i3, considering the urban mobility of this customer segment (high volume of low mileage journeys).

For those customers who are likely to drive over 5,000 miles per annum, we do offer FlexiMile Unlimited miles which provides uncapped mileage for the duration of the policy.  In addition to this, we provide the ability for customers to transfer from FlexiMile to FlexiMile Unlimited if their circumstances change during the period of cover.  This is done with no administration or cancellation charges to the customer.

It is important to note that the take-up that we have seen for FlexiMile Insurance (5,000 miles product) since launch would indicate that this is an attractive and valid product proposition.

  • Will this be offered in other countries?

Other countries are at this time investigating the feasibility of offering FlexiMile in their markets.  The UK is the only market to have FlexiMile as a live customer offering.

  • Is this something BMW wants to offer more broadly in the future and offer on other products of the BMW Group?

BMW Group Financial Services are investigating potential opportunities to extend FlexiMile across BMW & MINI, as well as developing Telematics offerings for the customers of BMW Group.

  • How does this fit in with BMW’s philosophy of offering “The Ultimate Driving Machine?” Is there a risk of penalizing our customers for enjoying their cars?

FlexiMile Insurance has been designed and developed with urban mobility in mind, specifically for i3, rather than combustion engine models.  Rather than penalizing customers, we are offering an insurance product suited to the urban mobility segment of the market which i3 is aligned to.

FlexiMile has been extremely popular with i3 customers, as evidenced by our 50% conversion rate from 7 day Complimentary Insurance to a FlexiMile policy.  To put this into context, our conversion on combustion engine models from 7 Day Complimentary Insurance to an annual policy is c11%.

It is also important to note that we operate in a market here in the UK where car insurance premiums are a significant cost in car ownership. Through FlexiMile Insurance, customers are seeing an average saving of 25% in premium when compared to what they would be charged on a standard comprehensive car insurance policy, without sacrificing cover.

July 25, 2014 18:30 rlanctot

Everyone tells me usage-based insurance (UBI) – car insurance based on your driving behavior – is already a huge market, rapidly growing, taking the industry by storm… take your pick. I am not buying it, although I am using it.

I participate in State Farm’s Drive Safe and Save Program in the U.S. The program requires the installation of an on-board diagnostic device in the OBDII port. (That’s my driving assessment appended to this report with the savings displayed above – with the “potential” savings available if I could just drive a little better.)

The companies that make the devices, the companies that process the data, and the carriers that provide the service all tell me there is monster growth in UBI. I believe them. Because what they are seeing is massive growth on a tiny base. So the growth will look very big indeed for the next few years.

But my own experience with usage-based insurance is that it is a little like sub-prime auto lending, which sometimes uses electronic modules on cars to enforce payment plans on credit-challenged customers. (These programs are usually referred to as “buy-here-pay-here” and include vehicle immobilizer technology to corral deadbeats.) It is a niche market with a significant "ick" factor.

UBI is best suited to drivers who simply can’t get a decent insurance rate without surrendering some privacy. In the UK, this is confined mainly though not exclusively to young drivers, for whom rates are exorbitant. UBI has also benefited in the UK from tagging along with anti-fraud devices and in Italy on anti-theft devices.

Every time I visit a client or speak at an industry event – including gatherings of State insurance commissioners - I ask the members of the audience if anyone is participating in a UBI offer. The answer is a universal “No!” So the industry that is promoting this technology is turning up its nose at the opportunity.

This needs to change if UBI is ever going to enter the mainstream. A friend of mine talks about “eating your own dog food” and that is precisely what needs to happen here. If UBI is to succeed, every broker and manager in the business needs to get a device on their vehicle right away.

Brokers need to understand:

1) How the devices are installed

2) What the data looks like

3) How the data does and does not change

4) How the data impacts the discounts

5) And what the discounts actually are

This is not happening in the industry today. My service is up for renewal with State Farm and my confused questions have gone unanswered at a variety of touch points. It is tempting, but not worth it to go into the details and, no, I did not record my conversation with the multiple agents with whom I spoke – tempting though it was.

The insurance industry wants everyone, as in EVERYONE, to be connected and tracked and evaluated. By tracking drivers insurers expect to be able to underwrite auto insurance more accurately.

Conventional wisdom in the industry is that adverse selection – ie. the rejection of customers for insurance coverage based on their electronically tracked driving behavior – will ultimately force all non-tracked customers into some sort of auto insurance purgatory. I say: “Nonsense.”

The most creative discount offers today are still coming from non-tracked offers that do not require any expensive hardware, software or Web portals. And too many customers simply can’t get past the “ick” factor inherent in tracking. It is a prospect killer – when it actually ought to be generating prospects.

What’s missing is trust and transparency. According to the pictured evaluation of my own driving, if I were to attain an A+ rating in all six driving categories I could more than double my discount. The only problem is…I have no idea what I am doing wrong and no means to find out.

Until insurance companies can find a way to instill trust with greater transparency and convince the brokers and agents to participate in the programs, UBI will remain a niche market and an insurance industry pipe dream. The creepiness of placing a tracking device on your car is hard to overcome, but the opportunity to develop more accurate risk models and underwriting schemes ought to be motivation enough to fix this.

Here is my recipe for UBI success:

A) Pay the customer to install the device

B) Install the device for the customer

C) Provide an on/off switch (within set time parameters)

D) Provide specific driving instructions as to how to improve driver ratings (and discounts)

Until these requirements are met, the UBI market will remain simultaneously huge and invisible.

July 24, 2014 16:08 rlanctot

Ford’s Mark Fields started 2014 with the announcement of aluminum-bodied F-150s and, following his appointment as CEO, has doubled down on the fuel efficiency message with the announcement of the hiring of Dr. Ken Washington, a top researcher out of Lockheed Martin’s space program.  Washington will head Ford’s advanced research and engineering efforts bringing experience in “light-weighting, (powertrain) control, autonomy and energy storage” to Ford, as highlighted in a brief interview with Strategy Analytics earlier this week.

The move coincided with a shift of Kumar Gulhotra, current vice president of engineering for Ford and Lincoln, to global president of Ford’s Lincoln luxury brand.

Washington was most recently vice president of the Space Technology Advanced Research & Development Laboratories, or STARLabs, at Lockheed Martin’s Space Systems company.  He is expected to accelerate innovation at Ford and will report to Raj Nair, group vice president, global product development.  Washington’s background spans nuclear engineering, information systems, supercomputing, information privacy and R&D regarding space-related technologies.

All indications suggest that the announcement revolves around Ford’s ability to meet Corporate Average Fuel Economy (CAFE) standards set by the U.S. government.  Washington’s boss, Nair, was quoted at the North American International Auto Show saying that Ford intends to double its global hybrid offerings by the end of the decade and expand auto start/stop to 70% of its cars.

Start/stop is already part of the plan for the 2015 F-150, shown earlier this week.  While Ford did not release its MPG estimates for the new F-150s, it did show a 732-lb. weight reduction from the use of aluminum, a surprisingly high figure.

The F-150 has always dragged down Ford’s corporate average fuel economy.  However, in January it claimed that the improved fuel economy of the new model will mean that for the first time the truck will help raise Ford’s overall fuel economy figure.  Ford has also had to restate the MPG of its hybrid models, as consumers complained of not being able to attain the originally claimed fuel economy levels.  So far, only the C-MAX, Focus and Fusion/MKZ have HEV/PHEV/EV versions.

It remains to be seen whether Lincoln will move toward a luxury HEV offering to take on Lexus or a more ambitious full EV to compete with Tesla.  The shift of Gulhotra and the hiring of Washington suggest that Ford is focused on fuel economy, electrification and autonomous driving, precisely in that order – even as it prepares to embed modems in its non-electric vehicles.

July 23, 2014 15:57 rlanctot

"So I think we are making some really good progress on the autopilot side, and I am confident that in less than a year you will be able to go from highway on ramps to highway exits without touching any controls." - Elon Musk, CEO, Tesla Motors

Consumers are clearly ambivalent about self-driving cars and electric propulsion. Multiple consumer surveys show significant resistance to autonomous vehicles, and enthusiasm for electric vehicles, aside from Tesla’s Model S, has been muted.

That’s why we need Tesla and Google. Like Henry Ford, Elon Musk does not make decisions based on consumer surveys, and Google pursues its own ideas about transportation and location awareness beyond the tyranny of the focus group or the legal department.

If left to its own devices, the auto industry will study electrification and autonomy to death on the test track without ever bringing an autonomous vehicle (and only a few electric vehicles) to the highway. Nissan, General Motors, Ford, Mercedes-Benz and BMW are all pushing hard at the concept of electric vehicles, but the progress has been incremental to date.

Tesla, meanwhile, is making the big bets on “gigafactories” and fast charging stations and shared patents. It is worth noting that BMW is trying to be a fast follower on the electrification front offering to share its battery technology Tesla-like. And Nissan is trying to match Tesla with the breadth of its charging station installations.

On the autonomous vehicle front, auto supplier Continental has highlighted the fact that legal and regulatory frameworks including the Vienna Convention and UN-ECE regulations continue to block the path of autonomous vehicle adoption by placing legal barriers in the way of precursor technologies such as traffic jam assist and emergency steering assist.

As reported in ITS International: “Article 8 of the 1968 Vienna Convention on Road Traffic specifies that the driver must maintain permanent control of the vehicle. This limitation was amended, however, in March 2014 in response to the increasing automation of vehicle systems. Automated systems are now permitted as long as they can be overridden or deactivated by the driver. This has established the legal foundation for partially automated driving since control of the vehicle may now essentially be assumed by systems as well.”

Aside from the fact that the Convention has not been ratified by all parties, the modification is absurd. In essence, it says automation is only allowed as long as it can be de-activated or overridden. This is entirely contrary to the mission of automated driving, which is to remove the 90% of accidents attributable to the driver, according to industry experts.

Further, this regulation focuses on so-called Level 3 automation (see definition below) under which the driver must always be prepared to take control. This is precisely the self-driving scenario that car makers are struggling with if not outright avoiding – and one that Google is embracing,

Also as reported by ITS International: “The United Nations Economic Commission for Europe (UN/ECE) offers regulation (UN-R 79:Uniform Provisions Concerning the Approval of Vehicles with regard to Steering Equipment). According to this regulation, automated steering is currently only permitted up to a speed of as high as 10 km/h.”

Once again, this regulatory authority is out of step with reality. Continental is correct. Both the Vienna Convention and UN-ECE regulations reflect a complete disregard for the facts on the ground.

In the highlighted quote at the beginning of this commentary Elon Musk, speaking at the latest Tesla shareholders meeting in response to a question says that Tesla believes drivers should be able to go from on-ramp to exit-ramp without touching the vehicle controls. Now, he did not say that Tesla will offer this capability, but what he did say was that this ought to be possible.

The most logical scenarios for automated driving will be parking, low-speed collision avoidance, and high speed highway driving. For its part Google is focusing on Level 3 autonomous driving scenarios which include all of the above and assume, at least for current Google cars, some driver participation.

As defined by the National Highway Traffic Safety Administration in the U.S., Level 3 Limited Self-driving Automation is defined as:

“Vehicles at this level of automation enable the driver to cede full control of all safety-critical functions under certain traffic or environmental conditions and in those conditions to rely heavily on the vehicle to monitor for changes in those conditions requiring transition back to driver control. The driver is expected to be available for occasional control, but with sufficiently comfortable transition time. The Google car is an example of limited self-driving automation.”

Car makers and their suppliers are leery of this Level 3 driving proposition, but Google is plunging forward. We need a Google to explore this space, because no level of market research, focus groups or number crunching will overcome the internal legal and liability objections of a typical car maker to explore Level 3 with cars on the road.

We need Tesla to prove out the concept of an attractive EV capable of being brought to mass market price points. Tesla is still working on this. In the words of Musk:

“The ground strategy at Tesla has always been a three step process where step one was little volume, high price product with the Roadster then we've got mid volume mid to high price car with the Model S and then generation three being the high volume lower cost car. And with Gen 3 with our third car, we're expecting that to be about a $35,000 car with a range in excess of 200 miles.

“But I should say that 35,000 may sound a little affordable, but when you consider the savings from use of electricity versus gasoline in the U.S. that translates effectively to like a gasoline car of about maybe $28,000 or in Europe gasoline car of about $22,000 to $24,000. So actually the affordability of it should be quite high, I am very optimistic about that and we're going to try to make that third generation car happen as soon as we possibly can.”

We need Google to prove out the Level 3 use case. Without the legal framework that Continental has requested, on behalf of the industry, getting Level 3 past the lawyers will be asking too much.