Big data, part 4: latching onto video, comparative analysis

Lytx’s new ActiveVision service integrates with in-vehicle technology and Lytx DriveCam sensors for an extensive level of data collection and analytics to provide in-cab and post-trip driver coaching.Lytx’s new ActiveVision service integrates with in-vehicle technology and Lytx DriveCam sensors for an extensive level of data collection and analytics to provide in-cab and post-trip driver coaching.

Editor’s note: This is the 4th and final part of a series on the use of big data technologies in trucking. For parts one, two and three click on the respective numbers. 

According to industry analyst firm Frost & Sullivan, video safety solutions have a meager 0.46 percent penetration rate in the fleet market, but they are growing quickly. This leaves a lot of room for more fleets to expand their big data capabilities.

Video has already proven to be one of the most effective tools that fleets are using to create behavioral change in drivers. The analytics behind video safety solutions add another dimension.

When a supervisor meets with a driver, “you don’t want to make things overly complex,” says Bryon Cook, vice president of data and analytics for Lytx. You simply show a driver a video of his or her bad behavior and discuss how to correct it, he says.

During the past 10 years, Lytx — originally called DriveCam — has collected more than 50 billion miles of driving data from fleets using its video safety system. Lytx also provides predictive and prescriptive information to fleet managers that show which driving behaviors are predictive of collisions and what are the risks of each.

Some of the findings may be counterintuitive.

“Not wearing a seatbelt is an indicator of accident risk, even though it doesn’t cause collisions,” he says.

Lytx continues to add new sources of driver data from sensors and external factors like traffic and weather to provide a more holistic view of drivers to make better risk assessments.

“A lot of information that we can learn about drivers comes from how they react to their environment,” he says. A future possibility of real-time analysis involves using its driver-facing camera to recognize facial patterns for fatigue, distraction and other risks. Rather than monitor the driver at all times, the facial recognition could be activated only when trends of risky behaviors are detected using other sources of driving data, he says.

The Safety Analytics dashboard shows drivers with the highest collision risk.The Safety Analytics dashboard shows drivers with the highest collision risk.

In addition to capturing video of risky driving events for review, PeopleNet recently added a new Safety Analytics dashboard as an option in its Video Intelligence application that encompasses information on violations of posted speed limits, hours of service, and risky driving behaviors. It also brings in daily CSA violation data from Vigillo to identify the most at-risk drivers in a fleet.

The dashboard is based on a four-tier scoring system that mirrors the crash predictors found in a 2011 study by the American Transportation Research Institute, says Jim Angel, vice president of video intelligence solutions at PeopleNet.

The dashboard identifies drivers by degree of risk. Those with scores in the top 10 percent in each category, and overall, are highlighted in red followed by yellow and green for the top performers.

 

Comparative analytics

Another ongoing advancement in big data and cloud computing platforms is the ability to compare results from one fleet to industry peer groups. This benchmarking ability makes it possible to set more realistic attainment goals.

Reporting tools in the SmartIQ Suite from SmartDrive compare fleet performance to industry peer groups.Reporting tools in the SmartIQ Suite from SmartDrive compare fleet performance to industry peer groups.

Lytx shows clients where they rank compared to peers in overall safety and fuel performance, Cook says.

“We typically do that with quarterly performance reviews,” he says.

SmartDrive Systems’ video safety platform captures and transforms complex event and telematics data into numerical values for every driver and for the overall fleet. These “SmartIQ Scores” use a numeric scale with a higher score indicating higher risk of collision, and are normalized by amount of hours and miles driven.

Fleets can use the scores internally to compare driving skills, risks, fuel performance and more for the different divisions, terminals, managers and other areas of a company. SmartDrive can also compare results from one fleet to the aggregated results of industry peer groups. As one example, a for-hire fleet could compare fuel efficiency on a specific lane to from Kansas City to Chicago, says Steve Mitgang, chief executive officer.

Another possibility is to compare the skills of drivers from different schools, he says.

In the near future, Angel says PeopleNet will develop benchmarking tools for the Safety Analytics dashboard for fleets to compare performance to peers with similar operational profiles.

Transportation companies that use one of the TMS platforms from TMW Systems are able to share data anonymously through a survey that collects data for industry benchmarking. TMW is working to automate the capture of key operating metrics directly from the databases of customers who choose to participate.

A few examples of industry data that TMW can provide include rate-per-mile between zip-code combinations. On the maintenance side, fleets can see the cost per mile of tractors and trailers by manufacturer, year, make and model. On the operational side, fleets can look at the average dwell time on a lane, revenue per hour calculation, the distribution of operating ratios, and more.

In the future, TMW plans to be able to capture data from fleets that use other TMS systems, as long as customers are willing to contribute to the big data platform, says Nick Orlando, director of business intelligence for TMW Systems.

Any TMW customer can access results through a web portal it calls a business assessment tool. As more data is added to the platform, Orlando says more metrics will become available for use in market intelligence.

Now that the era of big data has arrived, the evolution towards higher levels of predictive and prescriptive analysis continues, giving users at all levels in a company greater insights and direction to achieve results.