Predictive traffic is the single most important telematics application. This conclusion is borne out by the need for predictive modeling when creating navigation routes for truckers, taxi or bus drivers, consumers or emergency responders as well as the need for predictive modeling when creating traffic management systems for multi-jurisdiction metropolitan areas.
Around the world there are a number of interested parties that participate in or benefit from the creation of reliable traffic information. These interested parties include, but are not limited to, government(s), law enforcement, higher education, wireless carriers, mobile navigation suppliers, car makers, emergency responders and government contractors.
But no entity has more at stake in or will do more to determine the future of traffic information services than governmental organizations – primarily state and local. That influence is already having an impact on the prioritization and deployment of different data gathering systems based on their ability to accurately report traffic conditions.
Some of the earliest indicators are reflected in a recent presentation given by the New Jersey Department of Transportation as part of the Route 95 Corridor initiative. The NJDOT representative ranked traffic data sources for travel times based on their accuracy as follows:
1. License Plate Readers and Toll Tags
3. Inrix real-time
5. Cellular probe data
The NJDOT presentation reflects a high degree of frustration with cellular probe data, but it highlights the continuing dominance of infrastructure-based scanning and sensing technologies due to their accuracy. The larger implication is the endorsement of Bluetooth solutions, such as those from Trafficcast, Traffax, Post Oak, Alcatel-Lucent and Siemens, as a significantly lower cost infrastructure-based system for travel time measurement – an essential metric for traffic data systems.
Now take these implications one at a time. First, consider cost. Because Bluetooth scanners can be installed on existing infrastructure and deliver their information wirelessly, the cost of deploying these systems is a fraction (less than one third) of the cost of camera or RF-based technologies.
From an accuracy standpoint, Bluetooth scanners are capturing unique MAC addresses and have been shown to be comparable in accuracy to tolling and license-plate reading solutions (all based on point-to-point measurement) and superior to data feeds from suppliers such as Inrix, which are dependent primarily on vehicle mounted probes, and radar-based systems (which measure speeds) such as those deployed originally by Navteq’s former Traffic.com subsidiary – the media arm of which was recently spun out and reconstituted independently as Radiate.
Accuracy has emerged as the single most important criteria among government traffic information evaluators and, combined with the cost advantage, has thrust Bluetooth to the forefront of traffic data gathering tools. Bluetooth is also seen as more attractive from a privacy standpoint than RF technology (reading a tag assigned to a specific user) or cameras (scanning license plates or photographing drivers).
The other interested parties in the traffic eco-system are typically promoting different traffic data types correlated to their own personal, professional or commercial interests. Wireless carriers around the world have been sharing handset signaling data with third parties in the hopes it can be converted into useful traffic information. Mobile device makers have been plumbing their GPS probe data with the same objective in mind.
Governments and law enforcement have been deploying cameras, radar and RF sensors to capture traffic information as well as to enforce speed limits and collect tolls, respectively. Radar is a direct competitor and precursor to Bluetooth and has been promoted as a low-cost technology but with low data quality.
Universities are players in the discussion because they participate in research studies around the efficacy of different traffic information gathering strategies. The downside of university participation in the debate is the fact that some - such as Texas Transport Institute (Post Oak-Bluetooth), University of Maryland (Traffax-Bluetooth) or Massachusetts Institute of Technology (Vtrack-Handset probe) - have a commercial interest in one technology or another.
When it comes to determining which technologies will survive and thrive in the market, Bluetooth stands out thanks to the influence of governments. State and local governments command the power of the purse to fund the deployment and adoption of new or existing traffic information solutions.
Evidence of this influence is clear from Trafficmaster in the United Kingdom to Traffic.com and Inrix in the U.S. Each of these organizations was blessed at one time or another by public authorities seeking traffic data sources to support travel time calculations for variable message signs and other traveler services.
Bluetooth as a low cost and accurate solution arrives as government budgets are being pinched. The budget pinch has forced transportation departments to re-evaluate their data gathering strategies.
The budget and data re-evaluation has also given cause for transportation authorities to re-assess their data sharing models. Traffic.com, Trafficmaster and Inrix have all benefited financially from their government contracts. The government contracts have had the effect of immediate cash injections into these organizations along with the boost of a very public endorsement. These public authorities are looking for some payback.
Government support of traffic information data gathering has created the most reliable source of revenue for traffic companies. Other revenue sources – such as traffic information subscriptions – have been less reliable, especially in the context of available free sources of traffic information such as Google or even the municipalities themselves.
Better data and a piece of the pie
Now governments are seeking more accurate data – leading to the growing adoption of Bluetooth technology - and some are seeking revenue sharing agreements. While the shift to Bluetooth holds the promise of raising the level of accuracy of real-time and, indirectly, predictive traffic information, the pursuit of revenue shares with traffic information providers can have a detrimental effect on the industry.
It is true that at least two traffic information providers, Waze and Inrix, have received substantial equity injections suggesting extraordinary market valuations. And Traffic.com clearly benefited from early transportation contract wins for the deployment of its radar-based traffic sensors – which led to Navteq’s acquisition of the company for $170M. (In retrospect, what looks like an inflated price in the context of Navteq’s recent divestment of Traffic.com’s assets, will eventually look modest in the context of future traffic information provider valuations.)
The challenge facing the industry is simple. Traffic is the most important telematics application and consumers surveyed by Strategy Analytics routinely rank traffic information as their most important mobile device application. But reliable and accurate real-time and predictive traffic information remains elusive.
Car makers are seeking uniform, global traffic information solutions to support their navigation systems, but traffic data sources are inconsistent around the world. Even location referencing is not universal in spite of three standards (T-PEG, RDS-RMC, OpenLR) and one proprietary solution (Navteq).
Local governments are opting to create their own data fusion engines – a task normally performed by a third party – because they are dissatisfied with the available solutions. The minimal objective local governments are trying to resolve is to deliver travel times for traveler services. But urban area planning is also an issue.
With government organizations trying to do more with less available funds it is logical that they are turning to traffic information providers for a share of any resulting revenue. Unfortunately, this is undermining the commercial opportunity for these traffic information providers.
The nightmare scenario that may play out as a result of increasingly assertive government representatives is already playing out in China where municipal authorities charge exorbitant fees for their sensor data. Traffic data companies are struggling to make money and the accuracy and universal availability of data has been compromised. Consumers in China have been left with inferior traffic information, a potentially fatal flaw in their monetization plans.
Consumers take for granted the fact that traffic information is available worldwide, even if the accuracy is dubious. But the increasing cost of doing business could discourage market participants and retard the technological advances necessary to attract millions of traffic information subscribers.
The good news is that traffic information providers, outside of China, have been able to successfully resist the revenue sharing model. There appears to be a rational recognition that the traffic industry is caught in a chicken and egg dilemma. Under the current circumstances, profits will not flow until a more accurate and reliable product is made available.
Once consumers see that accurate and reliable traffic information is available, both real-time and predictive, they will be willing to pay. In the meantime, companies such as SiriusXM, Verizon Wireless and AT&T, which are aggressively pursuing navigation and traffic data subscription opportunities, will have to wait to see wider adoption of traffic information services.
SiriusXM is currently in the midst of its annual “shoot out” evaluation of traffic data providers. Attach rates for SiriusXM traffic data have typically been in the 10%-20% range reflecting both the unwillingness of many consumer to pay for this data and questions regarding its accuracy.
If SiriusXM is able to enhance the quality of its traffic data with a new, higher quality source the enhanced traffic service will go far toward anchoring SiriusXM’s data services and restore automotive OEM confidence. SiriusXM and its limping data services have, thus far, seen limited consumer adoption despite wide availability from North American OEMs.
But to achieve a new level of accuracy and reliability will require the deployment of new technology. Bluetooth is the newest and most promising candidate. Bluetooth installations are already in place in more than 30 states in the U.S. as well as in South America and Asia. And the participation of companies such as Trafficcast, Alcatel-Lucent and Siemens suggest a robust competitive environment – attractive to transportation authorities and auto makers.
In fact, Trafficcast has the first OEM award – from OnStar – which will benefit from Bluetooth data sources. Presumably, it will not be the last.
State and local governments around the world are calling the shots in the traffic information industry thanks to the power of their lately-diminished purse. Bluetooth technology is increasingly emerging as the preferred, accurate, low-cost, infrastructure-based solution for capturing travel time data, particularly in the U.S.
The accuracy of handset probe data is widely being called into question even as investment dollars continue to flow to handset probe data suppliers such as Decell, Cellint, Intellimec, AirSage, Waze, ITIS Holdings (now part of Inrix), TomTom and the Massachusetts Institute of Technology. Most of these parties have been asked to defend the accuracy of their data (by transportation authorities and this analyst) or provide test results validating accuracy claims. Handset probe data is especially poor in urban canyon type environments.
(A side note regarding accuracy testing: TomTom and Decell, among others, have used vehicle-based probes to validate handset probe data – GPS and cell signaling. Transportation executives are increasingly using data from fixed scanning technologies, such as Bluetooth, to validate handset probe inputs. This crucial change in data evaluation to a new definition of “ground truth” is undermining the credibility of the handset probe industry.)
The testing of state and local traffic authorities is a direct challenge to test and measurement protocols advocated by BMW - known as Qkz. Cell signaling technology used by Inrix's recently-acquired ITIS Holdings division has passed muster for use in Australia, Germany, Ireland, the U.K. and Singapore based on this testing protocol.
One barrier to global adoption of cell-signaling technology is the requirement of carrier cooperation. The other barrier is a robust technology that can be licensed and deployed across multiple networks, something ITIS has been able to achieve.
As governmental bodies develop their own data fusion solutions in the interest of delivering more accurate traffic and travel time information, drivers will increasingly receive public traffic information – either on variable message signs or via 511 and other traveler information services – that conflicts with private sources of traffic information. Additionally, some fragmenting of the traffic market may occur such that some urban areas have better traffic information than others.
All of this should be more reassuring than disturbing. The steady advance of traffic information gathering and fusion reflects the existing dissatisfaction with “good enough” traffic services. But it is transportation authorities that are raising the bar and paying the price for the kind of enhanced accuracy that will restore confidence to consumers and vigor to valuations. In time, revenue sharing may begin to make sense as consumers step forward and subscribe by the millions – establishing organic mass market demand for accurate and predictive traffic data.