When it comes to the safety of fleet drivers and vehicles, technology is advancing rapidly, bringing more sensors, more cameras, and more intelligence to increasingly connected vehicles.
Years of experience with basic telematics solutions have demonstrated how effectively they have enhanced safety, in addition to boosting efficiency, improving driving skills, and optimizing routes while managing insurance and liability exposure.
Both figuratively and literally, technology is now in the driver’s seat when it comes to safety.
Fleet safety technology is maturing beyond telematics with the use of technologies like video cameras and all-pervasive connectivity. Advanced analytics capabilities like artificial intelligence and deep learning have created essentially four levels of fleet safety solution maturity:
- Descriptive: This involves looking at certain historical factors reported on a weekly or monthly basis, and taking an action, often days or weeks after a critical event. Vehicle speed, for instance, rapid acceleration, harsh braking, or other aspects of a driver’s style. As a fleet manager, you look at driving data, see where there are problems, and coach drivers toward safer and more efficient practices. You can also generate in-vehicle alerts for speeding. Telematics systems have been doing these sorts of things for years.
- Predictive:These solutions take into account drivers’ history over a longer period of time, and other environmental factors such as weather conditions, ambient light, and neighborhood crime statistics and apply complex algorithms to predict the probability of different likely scenarios about what will happen next. These predictive scores can be used to train and guide drivers. As an example, you can leverage fleet data (many drivers in the fleet speed on one particular stretch of road) and individual data (Driver A is historically prone to speeding) to determine that Driver A will probably speed on that road as well. Knowing what to expect, you can take action to reduce the risk.
- Prescriptive:This type of solution leverages artificial intelligence (AI) and three to five years of historical data to prescribe corrective actions for the driver. This approach creates a feedback loop, in which you can monitor what the driver is being advised to do and whether the behavior has changed. Then the loop further monitors the effects of that behavioral change for additional guidance and improvement. Cameras and sensors, for instance, might combine to detect the light conditions outside and advise a driver to keep a greater distance between his or her vehicle and the one in front.
- Automated:With automated driving systems, there is a potential to reduce the number of serious crashes. Automated solutions use a combination of sensors, cameras, AI and machine learning, and high-speed network communications to improve driver safety.
While predictive analytics typically report on alternate risk outcome scenarios with varying degrees of confidence, prescriptiveanalytics is more precise and allows managers to prescribe actions. An algorithm might help prescribe specific training to a historically high-risk driver to address a particular behavior, then assess the effect of those prescriptions, in terms of whether the number of incidents declines after the corrective actions.
Another scenario might be in-cab alerts and guidance, such as advising a driver to slow down upon approaching a road curve, based on the system’s understanding of a driver’s regular behavior. In addition, such a system could also flag a driver when nearing a high-crime neighborhood. Taking crime data from the cloud, it could establish a “geo fence” of risk to the driver, the vehicle, or the cargo, and suggest a different route.
The smarter vehicle
With more and more sensors in vehicles, the amount of data is growing exponentially. This gives vehicles the opportunity to make intelligent decisions to improve driver safety. Smart edge devices can quickly generate alerts and notifications to help avoid critical accidents.
Today there are various sensor systems in a vehicle that don’t communicate much with each other, but vehicle systems of the future will interact with each other to improve the decision making. In fact, with advanced driver-assistance systems, a cooperative safety and efficiency ecosystem can be developed based on sensory inputs and advanced data processing from more than one vehicle. A whole fleet could function as a unit, sharing information and communicating with each other.
The vehicles could keep each other aware of traffic flows, weather changes, and other relevant data. This is all made possible by a macro network that assures broad coverage, high capacity, and low latency.
Cloud-connected vehicles have the benefit of advanced data and analytics-based applications. Fleet drivers can easily report unsafe and suspicious activities along with video recordings of an event. Using the cloud, fleet managers can track and manage vehicle safety across large fleets from a centralized web-based or mobile portal. It becomes easier to see trends and make better informed decisions.
More cameras, a lot more
Video technology is key to advancing fleet safety: dashboard cameras pointed forward, rear view cameras, side cameras, and even in-cab, dashboard-mounted cameras that monitor the driver.
Driver-focused cameras leverage AI to assess situations, such as a driver’s eyes diverting from the road for too long, or picking up a mobile phone or tablet. Or failing to use the seat belt. In a worst-case scenario, that could be a drowsy driver nodding off. Detecting such things, the camera system can alert the driver and even report distracted driving events to fleet managers, who can use it to coach drivers.
The camera’s recordings can also be uploaded for coaching and other purposes, which brings up the issue of driver privacy and the feeling of “being watched.”
These are concerns as we make increased use of in-vehicle video cameras. It’s a big reason why drivers tend to resist the improved technology, though their fleet managers are enthused about it because of the available data.
There are ways to overcome this, such as by not uploading every video feed generated on a given trip; perhaps only certain significant incidents or behaviors would warrant that. For lesser “offenses,” the system itself could just issue an alert to the driver.
But despite any resistance, we’re seeing that more and more drivers are recognizing the valuable benefits of this type of observation. Seeing that it can make them safer and better drivers and help them avoid risky situations, they are coming to embrace it and are influencing their fellow drivers.
That is especially true in instances where an accident occurs, and thanks to the onboard cameras, it can be proven that the wreck was not the fault of the fleet driver. Drivers appreciate that, and of course so do fleet owners.