By: Shuo Zhang
With the cruise control set to 50 mph, a driver navigates along a stretch of west coast highway. This driver would not likely be characterized as “reckless” by most people’s standards. But as an out-of-town motorist, he is unaware that he is going above the average speed that occurs on this roadway.
The posted speed limit of 55 mph provides him with nothing more than a false sense of security. As he decreases speed slightly to prepare to pass the vehicle in front, he is only about 1 mile away from a location that is notoriously high risk for accidents, with 100 collisions having already been reported this year, primarily due to the speed at which cars are merging from a connecting on-ramp.
At the same time, across the country where the winters are always harsher between the months of November and February, another driver is proceeding down a winding roadway that is generally safe most of the year, but is also known to be a place where many drivers accelerate more than they legally should.
On this day, however, a quick-moving storm front will bring a mix of snow and rain, rendering the driver prone to a crash, unless they are pulled over by the police, who are known to stakeout the location frequently, before a disaster ensues.
Both drivers should be able to benefit from better outcomes given the depths of data collection that governmental officials have at their disposal and the availability of global positioning system (GPS) technology today.
Intelligence, however, does not always equate to dissemination of knowledge among the public. Yet, two improved patents for a GPS program designed to be more “insightful” for those traveling America’s highways could soon improve the rates of injuries and death even if the government’s use of big data continues to be subpar.
Data Collection Conundrums
Traditionally, the potential for an accident could be chalked up to happenstance, coincidence, bad luck, careless driving, or a combination of these factors.
But in a time when the collection of data and analytics are seemingly a part of every facet of life, the potential for a highway accident can now be identified on a scientific level. Through the Internet of Things (IoT) and other systems, data can be harnessed to a point that predictive analytics could realistically aid any driver in any hypothetical (but real) situations to avoid tragedy—if they only knew risks that they were driving directly into.
Big data analysis is a mature technique that has been utilized by the U.S. government and other organizations throughout the country (and worldwide) in a variety of industries for many years to monitor public and community transportation patterns.
Yet, it has not been fully implemented in a way to assist drivers to commute more safely every day by preventing traffic accidents and citations. By categorizing traffic data into identifiable patterns or risks, roads could be constructed and navigated in more productive ways.
Data could be collected and analyzed in real time and, when combined with historical research, effectively predict potential for roadway incidents based on certain specifications and communicate potential risks to commuters through GPS user engagement panels featured on smartphones and other digital devices.
When this data is applied to a driver’s daily life, it decreases the possibility of accidents and even driving violations like speeding.
Why isn't this technology being implemented? The reasons are debatable. As easily as it could be assumed that the U.S. government does not want to invest significant dollars into such a program, it could just as easily be assumed that the financial wherewithal is there but a lack of data organization across the country is the main culprit.
Regardless, there is a prevailing sentiment that the public would have trust issues with government officials implementing big data, and potentially sensitive personal information, for this purpose.
Big Data Analysis
More than 38,000 people die in car accidents each year in the U.S. An additional 4.4 million people are injured severely enough that they require medical attention.
In 2020, despite the pandemic, car accident deaths in America spiked, with more than 42,000 people dying.
The National Highway Traffic Safety Administration (NHTSA) is tasked with enforcing vehicle performance standards and fostering partnerships with state and local governments to reduce motor vehicle deaths, injuries, and economic losses.
The NHTSA’s National Driver Register and Traffic Records Division provides coordinated accessibility of crash, roadway, citation, and other surveillance databases, and is tasked with helping states throughout the country to improve their respective traffic safety data collection, management, and analysis.
But there is not a consistent national standard for state and local governments to collect and store their data that enables the data to be consistently shareable or searchable. As such, officials and residents alike depend on the diligence of their respective states to track and utilize data effectively.
Some states do a better job than others. New York is an example of a state that does track its data well. The state’s DMV maintains statistical data about crashes from 1995 through 2014, and crash data from 2015 and beyond is publicly accessible, searchable, and archived by the Traffic Safety Statistical Repository (TSSR).
Predictive Analytics and Transportation Services
If the tracking of individual traffic accidents is the baseline of the requisite big data needed to improve driving safety, there are several ways that government officials could categorize traffic data into data types and user types that correspond to one another. Then, they could utilize predictive analytics to generate specific notifications about potential impending traffic accidents (and moving violations) based on the driver’s location and environment.
Examples of data sets that could be part of a comprehensive reservoir of information for tracking and mapping, as well as for developing purposeful predictive analytics include:
- Traffic accident-related data based on vehicle make and model
- Traffic accident-related data based on road type and condition
- Traffic accident-related data based on location
- Traffic accident-related data based on time of day/year
- Traffic accident-related data based on long-term climate and types of variable weather conditions
Improving Transportation Management
When drivers get into their cars today, they have multiple digital apps to choose from that will provide them with real-time traffic alerts throughout their commutes.
However, they use real-time data and more recent information as opposed to historical information. This begs the question: How valuable is it for a driver to know about an accident that has just happened a few miles away that was very uncharacteristic for that roadway but did not generate a notification to the driver? (It is probably worth acknowledging, however, that these very applications can contribute to today’s motor vehicle accidents when they are accessed in the act of driving, which goes against advisory guidelines and highway safety regulations.)
But these newly developed programs have been designed to provide GPS functionality that would offer increased services than today’s typical platforms, including:
- The potential for accidents and violations in specific location or along specific routes based not on real-time data but, rather, historical data
- Consideration of the multiple variable aspects such weather, road condition, and vehicle type
- Route suggestions based on less probability of accidents and violations as a means for helping more drivers to reach their destinations safely, with less emphasis on taking the “quickest” route
Additionally, the functionality will connect the mobile application to a server based on real-time information for drivers, such as weather conditions, and historical database information related to the location’s typical speed, violation/accident data, and regulations/traffic laws for purposes of comprehensive analysis that will be delivered to users with precise, short messages received prior to a commute.
Without such research, there are no patterns of knowledge as to why certain incidents occurred at certain locations under certain circumstances from an historical perspective.
Privacy Concerns
It is difficult to envision the issuance of federal guidelines for the consistent tracking of such intimate historical data based on individual drivers. Concerns about privacy abound, and the financial investment to support the technology, not to mention the money that could be lost on collections earned through citations, would be substantial and a deterrent.
It is possible for governmental officials to collect and disclose data anonymously to appropriate entities, which could result in an expert analysis for distribution and future benefit. As technology continues to be utilized by more people and impacts our lives repeatedly, exposure to such data will create more accurate and useful devices.
As more people use technology applications, receive relevant notifications, and impact the roadways by exposing reasons for incidents, the benefits of historical data are key.
In addition to buy-in from governmental agencies, acceptance needs to come from the public if the data and technology are truly going to sync for overall safety. Not everyone will be accepting of such information gathering. Some people will resist due to fears of data abuse, invasion of privacy, and/or security concerns such as data breaches.
Resistance to change is typical, especially when it comes to technology. But one could argue that is not a good enough reason for the government to ignore the potential benefits.
The earlier this type of data mining is accepted overall, the earlier the benefits would be realized. Removing personal, sensitive data for security purposes and establishing a “safe place” for specific analysis tools or methods to be applied would enhance citizens’ trust and quality of life.
If the information is used properly, the reduction in accidents and lives saved could be immeasurable.
Today’s roads could be safer to drive day to day, and with big data analysis available they should be safer. Everyday drivers can prevent potential traffic accidents, injuries, and fatalities if given access to the best data and technology. Data can be collected and analyzed in real time and through historical research. As more of these data are applied to drivers’ daily lives, the less possibility there is for accidents and violations to occur. But who is willing to financially support such a reality, and at what cost? The answer to this question could determine everything.
About The Author: Zhang has seven years of experience in the transportation industry, providing IT support for a variety of field transportation projects.
