Driver authentication: The key to accurate performance measurement

Until recently, video telematics had been focused primarily on one, albeit critical, use-case – be the high integrity system of record around accidents and other critical incidents, driver authentication and in most cases exonerate the driver and fleet, protecting them from liability and punitive damages. While this use-case is and will remain important, for modern video telematics solutions it has become the equivalent of table stakes – that is, not a point of differentiation anymore.

‍

The introduction of AI-driven driver coaching and analytics into the video telematics toolkit has moved video telematics from post-hoc analysis to preventive real-time driver alerts and provided the ability to build extremely accurate driver profiles based on the violations and near-misses that are a lead indicator of the risk of critical incidents. Rapid improvements in AI that have brought ADAS and DMS to fleet dash cams have solved a hard part of this puzzle. Driver authentication, or accurately establishing the identity of the individual driving a vehicle, is the other obvious part of this equation. The key functionalities in the RideView platform enabled by driver authentication include:

• Building accurate scorecards at both fleet and driver levels. Fleet leaderboards that show best performing drivers can form the basis of rewards and incentive programs, building driver engagement with the solution.
• Detailed driver profiles including most severe incidents, incident trends, and comparisons with average fleet performance
• Creation of effective coaching workflows using key events selected and annotated with comments and tags by both driver and fleet manager

‍

In the traditional scenario, finding out who the driver was when a critical incident happened did not need to be a scalable process, and in most instances, it was done manually by piecing together different data points – trip logs, driver rosters, and even manual analysis of video footage. Solving this reliably and at scale over hundreds of drivers across thousands of trips every day, on the other hand, is not trivial. Technology-driven solutions are the only option, and the RideView platform supports a number of modalities to let drivers authenticate and associate themselves with trips.

Authentication modalities:

• Mobile companion apps
• Smartphones and tablets have become ubiquitous in commercial vehicles, driven by multiple factors. In some instances, due to regulations like HOS(ELD), enterprise tablets have become commonplace in certain categories of fleets. In general, mirroring the broader population, most drivers carry their personal smartphones with them in the course of their regular work. Using these devices as a means for drivers to authenticate themselves is, therefore, an obvious option, and our companion SDKs for Android and iOS help our partners achieve that, using 2 different communication modes:
• BLE:
• Dashcams broadcast their credentials (typically IMEI) over Bluetooth, and companion apps in the vicinity can receive these credentials and prompt a driver to link to a specific device and send driver credentials back to the device and form an association for that trip. This process can be made more efficient on the application side by using heuristics like signal strength to automatically connect to the closest device, in a multi-device environment like a yard. In addition, for scenarios where a driver is associated primarily with one vehicle, we provide a driver persistence mode, where credentials once set can be used for subsequent trips. Bluetooth has the added advantage of being a common modality across both iOS and Android devices.
• Wi-Fi Direct:
• An Android-based dash cam and an Android smartphone or tablet can set up a WiFi-Direct peer-to-peer network to exchange information. The dash cam broadcasts its credentials, and similar to BLE, a companion app using our SDK can prompt a driver to connect to a specific device. On Android devices, a Wi-Fi direct connection has the added advantage of being able to send larger sizes of data back and forth, and support a wider variety of use-cases including installation and diagnostics.

• NFC
• Near-Field-Communication-based smart cards are already quite popular as electronic identity documents, and for dash cams that support NFC, this is an option that we support as well. Tapping a smart card programmed with driver credentials onto a dash cam prior to starting a trip can be made a part of the driver’s pre-trip checklist. Most modern smartphones are capable of NFC card emulation, so the same workflow can be implemented with existing driver smartphones as well.
• Face ID
  
• All of the above steps need some manual intervention on the part of the driver, however minimal. Facial recognition accuracies have made significant leaps of improvement driven by recent innovations in AI, and hold the promise for a completely intervention-less driver authentication process, especially for devices with a cabin facing camera. The RideView platform recently added support for face-based driver authentication, supporting the following workflow:
• Fleet managers enrol a set of images for each driver, which are used to generate a learned representation for each driver that is stored in the cloud
• During a trip, appropriate images from the driver-facing camera are automatically selected based on the pose of the driver and uploaded to the cloud
• A matching algorithm extracts the representation from the uploaded images and reports the closest match within the set of enrolled drivers, or the failure to find a match
• Trips are automatically tagged with the driver ID corresponding to the matched driver
• Additionally, for incorrect matches, fleet managers are given the option to send feedback/correct the assignment

‍

Even outside of video telematics, accurate driver authentication is key to other telematics applications like HOS, creating an opportunity for TSPs to build a unified infrastructure for authenticating drivers across multiple applications. Where possible, we encourage our partners to leverage their existing authentication infrastructure to work with our platform, while utilizing the tools we provide in the form of companion apps and NFC smart cards. In other cases, as in Face ID, where we are bringing an entirely different modality to the authentication toolkit, we are happy to work with them to extend its uses to other use-cases even beyond video telematics. Video telematics sits within an ecosystem that includes other critical fleet applications, and we are committed to doing our part in improving the overall user experience for end-users.

‍

To find out more about driver authentication on the RideView platform, write to us at info@lightmetrics.co.