INRIX's Moment of Truth
“Truth is a pathless land.” – Jiddu Krishnamurti
Car companies and their suppliers are increasingly seeking to deliver a contextualized experience bringing together traffic, weather and parking information to assist drivers with real-time and predictive data. The challenge in creating and delivering these services is determining something called “ground truth.”
Ground truth is the real-time status of traffic, parking availability or weather or all three. We know we have multiple good sources for all of this information, but we also know we don’t have perfect information – ground truth.
For a car maker, the objective is to align with the partner or partners whose information is least wrong. There are a couple of tools – Q-Bench and Qkz - for measuring traffic accuracy (ie. ground truth), though few organizations outside of BMW actually apply these tools. Most traffic studies are conducted in an ad hoc manner with cameras, stopwatches and cars.
Weather forecasting continues to improve, but car makers and governments are looking for mesoscale weather down to square kilometers or even to the street level. That remains a work in progress.
Which brings me to parking - the thorniest rat’s nest of data accuracy to unravel. With dozens of variables, thousands of garages and hundreds of thousands if not millions of spaces to track, the parking business has become a game of probabilities rather than a business of exactitude. Getting parking information correct is more important than ever as cities increasingly seek to restrict the amount of available parking in order to discourage drivers.
Application service providers in the parking business have rapidly discovered that it is impossible to simultaneously know the availability of all parking spaces in any city at any given time along with all the correlated site attributes. So, again, the challenge is to be the least wrong service provider.
Why is parking so hard? Parking garages come in a wide range of shapes, sizes and locations. They have different parking rates, opening hours, access points, available space sizes, vehicle clearances, parking restrictions and on and on. To make matters worse, any number of these or a couple of other dozen variables may change from time to time or day to day.
INRIX, which bought ParkMe last year, attempted earlier this year to claim the superiority of its parking app to Parkopedia in Webinars, press releases and other announcements. GPSBusinessNews weighed in on the subject: https://tinyurl.com/hkdb6rv - “INRIX Claims Better Parking Data Than Parkopedia But Study Non Conclusive.” Parkopedia definitively rejected the INRIX findings.
The challenge for parking app/service providers is that it’s simply not possible to gather, process, deliver and maintain all of the data. But the implied warranty to the consumer is that the app or service is comprehensive.
This expectation is shared by car makers which themselves must conduct their own ground truth studies which rarely agree and usually overlook basic pitfalls in the data including inaccurate or out-of-date signage, restricted parking spaces and dynamic variables such as fees and opening hours. One key factor complicating the ground truth assessment published by INRIX is the distinction between public information on the Parkopedia Website and the data and factors embedded within the application.
In summary, the potential sources of data variability include data source (online or in the app, accessible to consumers vs. accessible to embedded application users), garages selected, time of day, city, and variables selected.
INRIX has also been touting some of its own selected and flattering ground truth data in traffic pitches to OEMs seeking to undermine competitors HERE and TomTom. It’s worth noting that Waze is increasingly included in this competitive set by all three parties.
The real challenge for INRIX is the fact that any perceived superiority of its traffic data from the recent past has more or less evaporated in Europe. In the U.S., some believe a rough parity has been achieved between the top three providers with TomTom and HERE surging ahead in accuracy over the past few years closing any “perceived” ground truth gap with INRIX.
All of this is important as INRIX faces a real moment of truth with its most strategic traffic data account: BMW. Senior INRIX executives have long regarded the BMW relationship as the linchpin to the company’s global traffic data strategy. BMW’s business must be maintained at all costs.
INRIX has watched as its data was substituted with TomTom flow data by Total Traffic and Weather Network at Toyota and in legacy BMW systems in the U.S. (with the exception of BMW’s ARTTi traffic in the 7 Series). In Europe, Volkswagen, Daimler and Audi all opted for TomTom - with Audi’s switch from INRIX to TomTom for its “XDTraffic” solution perhaps being the most painful for INRIX.
The most likely outcome is a half-a-loaf decision from BMW ceding Europe to TomTom and retaining INRIX in the U.S. If this comes to pass, it leaves INRIX with Porsche, Volvo and Tesla Motors for traffic (Toyota in Europe) – all representing tiny volumes ill-suited to INRIX’s vision of aggregating vehicle probe data for enhancing its traffic information. But hanging on to BMW, even if only for the U.S., may well represent a company-saving moment.
If TomTom has won BMW’s business in Europe, as industry sources suggest, it is likely to spark some agonizing strategy reappraisals at HERE, partially owned by BMW. In an ideal world, the consortium owners of HERE – BMW, Audi and Daimler – ought to be seeking ways to aggregate their data. It appears the aggregation of this valuable dynamic information will have to wait.
It is a good time for INRIX and the rest of us to ponder the teachings of Krishnamurti. Speaking on August 3, 1929, the day he became the leader of the Order of the Star in the East (an organization founded in 1911 to proclaim the coming of the World Teacher) he dissolved the Order at an annual member gathering saying in part:
“I maintain that Truth is a pathless land, and you cannot approach it by any path whatsoever, by any religion, by any sect. That is my point of view, and I adhere to that absolutely and unconditionally. Truth, being limitless, unconditioned, unapproachable by any path whatsoever, cannot be organized; nor should any organization be formed to lead or to coerce people along any particular path.”
I happened upon Krishnamurti while reading Ben Kelley’s estimable “Death by Rental Car.” The book recounts the fatal car crash that took the lives of the two daughters of Cally Houck. As a side note Kelley describes Cally’s study of Krishnamurti’s teachings. In the automotive industry different rules of thumb, protocols, prejudices, and thought processes dictate auto maker decision making processes in the same way that we pursue a spiritual path to truth.
These differing protocols, prejudices and thought processes ultimately govern how ground truth is determined for traffic and parking information. No two studies by two different auto makers will produce the same results. The most unfortunate thing is that the information providers typically are not allowed to influence these studies, even though the information providers are more familiar with potential sources of deceptive outcomes in the data.
It’s possible that some measure of peace may one day be achieved in the ongoing parking and traffic arms race. Today’s ground truth is that a growing number of traffic information providers are cropping up with new technologies (ie. Trafficcast: roadside Bluetooth and Wi-Fi censors) or hyper-local solutions for on- and off-street parking. With each new service comes a new ground truth.
It’s likely that only time and experience will determine the ultimate superiority of any single app or service overall or for any particular provider in particular local markets or for particular circumstances. Until there are widely agreed on standards – which will surely be shown at one time or another to have their own shortcomings – true ground truth will remain elusive.
Under these circumstances it is best to hue to Krishnamurti’s guidance and thought leadership. In the word’s of one biography (www.jkrishnamurti.org):
“He reminded his listeners again and again that we are all human beings first and not Hindus, Muslims or Christians, that we are like the rest of humanity and are not different from one another. He asked that we tread lightly on this earth without destroying ourselves or the environment. He communicated to his listeners a deep sense of respect for nature.”
Unlike Krishnamurti we are compelled to continue our pursuit of ground truth. May we recognize it when we find it.
Roger C. Lanctot is Associate Director in the Global Automotive Practice at Strategy Analytics. More details about Strategy Analytics can be found here: https://www.strategyanalytics.com/access-services/automotive#.VuGdXfkrKUk
Safety, Assets, and Innovations Engineer at WSDOT
7 年Food for thought. - Ted Bailey, P.E.
Technology Transformation Leader | Enabling Digital Innovation in Retail, CPG & QSR at Scale
8 年Good read Roger, Of-late, I am following every post of yours. Thanks a lot for sharing your knowledge and wisdom. Keep writing, Cheers, Raghu
Automotive Director at Mediamobile
8 年Spot on Roger - love to read your posts when you are "walking lightly on this earth without destroying..." :)
Shaping the fuzzy front end of Systems Engineering.
8 年Roger. Indeed quality assessment is a thorny issue, and getting good ground truth data even for (statistical) assessment often difficult.. Nonetheless tools like QKZ and QBench have long spread out in the ITS industry, well beyond BMW only. TISA (www.tisa.org) views that quality assessment must focus on the quality of the end-user experience on the road TISA has 'standardardized' TISA QBench for the benefit of its members and beyond. The TISA-Qbench specifications are available to the general public at https://tisa.org/newsroom/supplementar/