While redesigning our coaching tool, we wanted to deeply understand the place of driver coaching in our fleets’ safety programs:
- How much time do safety and fleet managers allocate to it?
- What do their ideal workflows look like?
- Which types of coachable events are most effective in improving driver behavior?
- Which types of coachable events resonate most with drivers?
With Nauto’s Collision Avoidance alerts, most of the driver coaching is done without a manager's intervention. In fact, Nauto’s Driver Behavior Alerts reduce distracted driving in 4 out of 5 drivers — without manager intervention. However, fleets still need to leverage Manager-Led Coaching for drivers that need additional training.
We learned how little time fleet managers can dedicate to 1-on-1 coaching. Even with dedicated coaches, each division or branch manager must manage many drivers. In any case, managers only have a short period of time for coaching sessions with their drivers.
To fit within this workflow, we wanted to build tools and capabilities to clearly prioritize drivers that need immediate attention, and help safety and fleet managers hold as effective, efficient and productive coaching sessions as possible.
Identifying High-Risk Drivers
Manager-Led Coaching helps safety and fleet managers identify drivers that need additional training using our proprietary VERA Score®. For each driver, the AI-enabled product filters highly selective, coachable events. This enables coaches to maximize their time and the behavior change impact on the driver. Finally, coaches need to be able to document, save, and export notes from each session, whether for a driver’s record or liability reasons. This is all built-in to the workflow.
AI-Enabled Coachable Events
To filter a driver’s riskiest events, traditional solutions often use a g-force based threshold. Every event above that threshold (braking, acceleration, collision...) is sent to coaches for review. This is an extremely narrow approach, and often overwhelms users with hundreds of events a week to address. In this context, coaching loses its effectiveness, and managers may not know how to prioritize their coaching efforts.
At Nauto, we leverage our artificial intelligence (AI) capabilities to find the most relevant coachable events. Multiple inputs feed into the curation of coachable events:
- First, we built and trained an AI model to analyze the sensor signature of the accelerometer and identify when a driver encounters a situation they were not anticipating. This is often a reliable indicator that a driver has missed a key environmental cue that would have allowed them to anticipate such situations.
- To confirm this, our human-review team manually assesses and extensively labels each event.
- Finally, each event is bucketed into different focus areas that the driver can be coached on. We surface the top 3 focus areas in the coaching dashboard, so coaches can focus on recurring, risky behaviors.
In addition to these AI-detected events, we heard from coaches that collision and near-collisions are usually worth a conversation, even when the driver is not at fault. We include these as coachable events. Finally, we’re also highlighting drivers with policy violations (no seat belt, cellphone, smoking) as well as high-severity distracted driving events — the most common cause of fleet collisions.
The list of focus areas will expand in the coming months to become more exhaustive and granular. At launch, focus areas will include:
- Distracted driving
- Drowsiness
- Road awareness
- Tailgating
- Traffic violations
- Aggressive driving
- Risky maneuvers
- Near-collisions
- Collisions
- Device tampering
Documenting Coaching Sessions
With the number of drivers each branch manager is responsible for, we understand how much a streamlined documentation tool can help safety and fleet managers assess their coaching effectiveness and track a driver’s progress. Saving notes after reviewing events with drivers is now part of the workflow. All coaching notes are kept indefinitely to provide managers with the entire coaching history of a driver. Users can also export these notes along with the details of each session.
And We’re Not Done Yet
In the coming months, we’ll continue to expand the capabilities within Manager-Led Coaching to track and measure coaching sessions and effectiveness across fleets and subfleets. This update will give visibility to executives and senior leadership into their coaches’ performance, and will track a driver’s behavior improvement progress after each coaching session.
Finally, we understand that some fleets don’t have time, personnel or suitable environments to hold 1-on-1 coaching sessions. For these situations, we are working on remote-coaching capabilities that will enable managers in the field and their drivers to seamlessly manage coaching via mobile tools.