Fleet owners and managers know there are good days and bad days for their drivers. A new report from Lytx, a company that specializes in machine vision and AI (artificial intelligence) applications for commercial fleets, shows what those good days and bad days are and what can be done to improve the odds of making it through those bad ones.
Lytx’s second annual “State of the Data” presentation at the American Trucking Assn. Management Conference and Exhibition was based on more than 6.2-billion miles of driving data from Lytx trucking clients this year alone. The presentation highlights noteworthy trends in truck-driving safety and concentrated areas of risk.
With one of the fastest-growing databases of analyst-reviewed videos of commercial driving behaviors, Lytx is uniquely positioned to detect driving and roadway risks, as well as to identify trends and opportunities that can help save lives and improve the safety of the communities. Lytx captured 5.5 million risky events between November 2018 to August 2019 from its trucking clients.
Lytx has found the state of Pennsylvania to be the second riskiest in the country, after Texas, for trucking fleets in terms of number of incidents captured there. According to Lytx’s data, the average severity of incidents captured in Pennsylvania was twice as high as that in Texas, and Pennsylvania was also found to have the most severe collisions of any state.
The road segments that contained the highest concentration of risk, according to Lytx’s proprietary risk-scoring system were:
- Pennsylvania Route 309 intersection with Pennsylvania Route 145 and E Rock Road (Allentown, Pa.)
- Pennsylvania Route 309: East of W Emaus Avenue (Allentown, Pa.)
- Interstate 85 intersection with University Station Road (Hillsborough, N.C.)
- Interstate 84 intersection with Pennsylvania Route 435 (Dunmore, Pa.)
- Interstate 81 / Pennsylvania Route 309 intersection with E Northampton Street (Wilkes-Barre Township, Pa.)
Areas of Pennsylvania, and particularly Pennsylvania Route 309, continue to dominate in terms of risk, holding onto the title of first and second riskiest roads in the country for the second year in a row.
All five of the identified road segments, including those on Pennsylvania Route 309, are near interchanges or on/off ramps, which naturally represent areas of increased risk. The sudden lane changes and rapid changes in driving speed associated with these areas tend to amplify risk, especially amongst already risky drivers.
When Lytx expanded its view from single-sq.-mile road segments to 60-by-60-sq.-mile areas, it found that trucking fleets displayed the most risk throughout greater Chicago. As a major transportation hub, and a common stop on western, eastern, and midwestern routes, the area surrounding Chicago was found to be 18% more risky among trucking fleets than the next riskiest area.
Riskiest TimesLytx data uncovered a shift in the riskiest days of the week and times of day from last year’s analysis. Between January-August 2019, Lytx identified:
|Day of week with most collisions||Thursday||Wednesday|
|Time of day with most collisions||Morning (5:00 a.m. – noon)||Overnight (11:00 p.m. – 5:00 a.m.)|
|Day of week with most near collisions||Wednesday||Friday|
|Time of day with most near collisions||Morning (5:00 a.m. – noon)||Afternoon (12:00 – 5:00 p.m.)|
|Day of week with least collisions||Monday||Monday|
Lytx trucking clients experienced 13% fewer risky driving events from January-August 2019 compared to the same period last year. The data shows improvements in several high-risk driving behaviors between November 2018 and August 2019, with drowsy driving as the most improved behavior:
|Most Improved Behaviors||
Decrease from 2018-2019
While the relative prevalence of late response has decreased from last year, this behavior remains an issue in trucking. Late response, which Lytx defines as an event in which the driver was not known to be distracted yet responded late or abruptly to a readily visible risky situation ahead, occurs 138% more often in trucking than any other industry. The same is true for not wearing a seatbelt; Lytx observed unbelted drivers 74% more frequently among trucking clients than any other group.
When compared to drivers in Lytx’s other client segments, including waste, distribution, transit, government, field services, construction, and concrete, there are several high-risk behaviors truck drivers avoided more than drivers in the other segments:
|Loose object in the cab||73% less often|
|Aggressive||38% less often|
|Blank Stare (day dreaming)||17% less often|
Aggressive driving occurs when the driver exhibited unsafe and/or unlawful actions, such as tailgating, weaving through traffic or excessive speeding, showing a disregard for their own safety, other drivers, pedestrians, or property. Blank stare occurred when a driver’s eyes were looking forward, but there was very little to no eye movement or activity, and as a result, the driver responded late to a situation ahead. Commonly referred to as daydreaming, this is one of the behaviors most correlated to getting into a collision. In fact, drivers who exhibit blank stare are 200% more likely to experience a collision in the next 90 days than a driver who does not display the behavior.
There are other risky behaviors the industry will need to address. Lytx’s data shows the most prevalent behaviors among truck fleets and year-over-year frequency comparison:
Top 5 Behaviors (in order of prevalence)
|Driver not wearing a seatbelt||Increasing|
|Cell phone use||Increasing|
|Following distance: ≥ 1 second to < 2 seconds||Decreasing|
Given the unique operating characteristics of large trucks, they have longer stopping distances than other vehicles. This makes late response a critical behavior to identify. Drivers sometimes have a false sense of safety and security in large trucks due to vehicle size and may fail to recognize the importance of seat belt use, especially in rollover crashes.
About the Data
Insights are derived from Lytx’s trucking-industry client database. The data is anonymized, normalized and in instances of concentrated areas and times of risk, generalizable to the trucking industry.
– with reporting by Corey Micho
Want to tweet about this article? Use hashtags #construction #IoT #infrastructure #AI #artificialintelligence #machinelearning #bigdata #digitaltransformation #cybersecurity #blockchain #Futureofwork