The addition of MV + AI capabilities to Vehicle Event Data Recorders (VEDR) is a huge step forward for independent service providers’ ability to address safety concerns, reduce liabilities, and improve overall operational efficiency. So, what exactly is MV + AI?
MV = Machine Vision
Machine Vision enables the device to visually monitor both the roadway and the cabin. Essentially, it serves as the ‘eyes’ of the camera.
AI = Artificial Intelligence
Artificial Intelligence allows the device to detect unsafe driving events. In other words, it’s the ‘brain’ of the camera.
MV+AI allows VEDR devices to detect specific patterns or conditions that indicate unsafe driving behaviors so they can be corrected in real-time, before an accident occurs. By understanding and responding to these unsafe events, fleets can significantly enhance their safety performance and foster a culture of safety with continuous improvement.
How are events recognized with MV + AI VEDR?
The MV + AI VEDR system proactively monitors and improves driver safety by triggering alerts and creating events based on unsafe driving behaviors. Unlike previous VEDR functionality, it does not require a g-force impact to register these behaviors.
Let’s take a look at this video example of how recognition looks in real-time:
The VEDR system automatically detects when a driver exhibits unsafe driving behavior such as speeding, distracted driving, or not wearing a seatbelt. In most cases, the driver is then issued an audible alert and given a brief time window to correct the behavior. If corrected, the alerts stop, and no coachable event is created. If not corrected, the driver will receive a second alert and once again have time to self-correct. If corrected, the alerts stop. If not corrected, a final alert is issued indicating that a coachable event has been logged. It will then go through a human review and ultimately end up with the driver’s fleet management team for further training and coaching.
Which types of events are recognized by the MV + AI VEDR system?
The MV + AI VEDR system is designed to monitor a wide array of driving behaviors derived from Key Indicators of safety. Here’s a detailed look at the AI Events the system recognizes:
1. Headway Monitoring Warning (Close Following)
Description: This event is triggered when the vehicle is following another vehicle too closely, with a time gap of less than three seconds. Close following is a common cause of rear-end collisions, and this event helps prevent such incidents by alerting the driver to increase the following distance.
2. Headway Monitoring Emergency (Tailgating)
Description: An emergency version of the headway monitoring event, this is triggered when the time gap between vehicles is less than one second. Tailgating significantly increases the risk of collisions, and the system issues an urgent alert to the driver to back off and maintain a safer distance.
3. Forward Collision Warning
Description: This event is activated when the system detects a rapid approach to a stopped vehicle ahead. The forward collision warning is crucial for preventing high-speed impacts, especially in heavy traffic and at intersections.
4. Cyclist Collision Warning
Description: Cyclists and motorcyclists are particularly vulnerable on the road. This event is triggered when the system predicts a collision with a cyclist or motorcyclist within a three-second window. The alert provides the driver with critical time to adjust their driving and avoid a potential accident.
5. Pedestrian Collision Warning
Description: Similarly to the cyclist collision warning, this event is activated when the system detects a pedestrian in the vehicle’s path, with a potential collision predicted within three seconds.
6. Lane Departure Warning
Description: Lane departure events are triggered when the vehicle crosses solid or double solid lines. This can indicate distracted driving or drowsiness, both of which are major risk factors for accidents. The system alerts the driver to correct their course and stay within the lane.
7. Stop Sign Violation
Description: This event occurs when the vehicle fails to come to a complete stop at an intersection with a stop sign. Rolling stops are a common traffic violation that can lead to dangerous accidents, especially at busy intersections. The system reinforces the expectation that the driver comes to a complete stop before proceeding.
8. Speeding
Description: Speeding events are triggered when the vehicle exceeds the posted speed limit by more than 10 mph. Speeding is a leading cause of accidents, and this alert encourages the driver to slow down to safe and legal speeds.
9. Over Max Speed
Description: For line-haul operations, this event is triggered when the vehicle exceeds 75 mph, regardless of the posted speed limit. This threshold is set to ensure that long-haul drivers maintain a safe speed, particularly on highways where high speeds can lead to severe accidents.
10. Using Phone
Description: Distracted driving is a major safety concern, and this event is triggered when the system detects that the driver is using an electronic device while the vehicle is in motion. This includes holding a phone up to the ear or frequently looking down at a device. The alert prompts the driver to focus on the road.
11. Distracted Driving
Description: Beyond phone usage, this event covers general distractions, such as frequently looking away from the road or engaging in activities that take the driver’s attention off driving. The system alerts the driver to refocus on driving to prevent accidents caused by inattention.
12. No Seatbelt
Description: This event is triggered when the system detects that the driver is not wearing a seatbelt while the vehicle is in motion. Seatbelt use is one of the most basic and effective safety measures, and this alert ensures compliance with this crucial safety practice.
13. No Driver Detected
Description: This event occurs when the system does not detect the driver’s face while the vehicle is in motion. This can indicate that the driver has left their seat or that the camera is obstructed. The system alerts to this condition to ensure that the vehicle is being operated safely.
14. Drowsiness Detection
Description: Drowsy driving is a significant risk factor for accidents. This event is triggered by a combination of factors, including yawning, head nodding, changes in eye gaze, and blink rate, analyzed over a rolling 30-second window.
Using AI Events to Build a Culture of Safety
With AI providing real-time feedback and opportunities for self-correction, MV + AI VEDR can help drivers avoid preventable accidents while developing safe driving habits. This also gives management improved insights into the driving habits of their team, allowing them to formulate impactful safety conversations and programs tailored to their specific needs.
Over time, this leads to a stronger safety culture where drivers are not just complying with safety protocols but actively participating in maintaining and improving the safety culture of their organizations. That can have a substantial impact.
At Descartes GroundCloud, we have seen our customers achieve 11% fewer claims’ costs compared to industry averages when they utilize the safety functionalities of VEDR and Proactive Safety. That translates to fewer accidents, less time spent dealing with claims, improved CSA scores, and a better bottom line overall for their businesses. It all leads back to making safety an organizational priority from the top-down.