Using Machine Vision and Artificial Intelligence for a Safer, More Efficient Fleet

By: Olivier Pierini - Director of Product Marketing, Lytx

From unpredictable traffic conditions to bad weather, from locked delivery gates to unforeseen construction delays, drivers deal with dozens of potential operational and safety risks every day. Any of these can cause collisions or lawsuits, reduce your fleet’s efficiency, and cost time and money. Identifying potential risks is essential to taking steps to reduce them for a safe, productive, and effective fleet.

There are hazards on the road and in the vehicle that traditional telematics sensors alone simply can’t solve. The power of video with machine vision and artificial intelligence (MV+AI) on top of traditional telematics sensors provides an expanded view of risk to help reduce transportation’s toughest problems like distracted and inattentive driving.

Get a bigger picture with MV+AI-triggered video

Essentially, MV+AI works as the eyes and brain of a dash cam. Machine vision sees and identifies driving behaviors. Artificial intelligence uses computer algorithms to interpret what machine vision sees.

Lytx’s DriveCam® device automatically records driving events that can be detected by accelerometers, such as cornering or hard braking. But other events with a low or no G-force, like rolling a stop sign, are not detected by an accelerometer. That’s where video and MV+AI can fill in the gaps.

Lytx’s dash cams use machine vision to recognize behaviors and artificial intelligence to detect patterns. It can identify in-cab events like when a driver is unbelted, smoking, food and drink, mobile device usage, inattentive driver, and an obstructed lens, along with road events like lane departure, following distance, critical distance, and rolling through a stop sign.

When these things occur, MV+AI triggers an audio alert to encourage a driver to correct their behavior in the moment*. A recorded event can also be sent through the cloud where managers can access and coach the driver on it later.

For example, suppose one of your drivers frequently rolls through stop signs in residential neighborhoods. The Lytx MV+AI trigger for rolling stop sign will detect and send events back to the fleet manager. Now you can coach the behavior before it results in traffic tickets, collisions, or damage to your brand.

Of course, video of quick-thinking defensive driving can be used to positively recognize your drivers, too—which is an excellent tool for driver morale and retention.

Why choose Lytx for your fleet safety challenges

Not all MV+AI is the same. The depth and quality of data is paramount to ensuring accuracy and eliminating false positives or missing significant events. Lytx® is an innovator in machine vision and artificial intelligence technologies, with over 25 years’ experience helping fleets of all sizes improve their safety and operations. At Lytx, algorithms are trained to identify risk using more than 250 billion miles of commercial driving that have been reviewed and tagged by professional reviewers. The Lytx professional event review team evaluates 10 million events every month and tags them with behaviors that can be used to train the MV+AI algorithms.

With over 145 million events and over 55 billion minutes of video from all over the world, spanning dozens of vehicle types, many different road conditions, and almost every weather condition conceivable, our MV+AI technology is built on the world’s largest database of commercial vehicle driving data of its kind. Highly accurate MV+AI algorithms depend on high quality data, and Lytx is a leader in the industry when it comes to the breadth and depth of in-cab scenarios available for algorithm development and training.

Not only is the quality of the data important, but Lytx dash cams have the robust processing power needed to take advantage of MV+AI. Lytx’s SF-Series cloud-connected event recorders are designed to deliver the MV+AI-powered insights you need quickly. By combining real-time processing and response with the power of the cloud where data can be aggregated and analyzed, you can address the most urgent issues now and go back for more information later.

Beyond our MV+AI triggers, we continuously develop innovative new features to help ensure fleet safety. Our Parked Highway/Ramp feature is one example—an event is triggered when the DriveCam® Event Recorder identifies that a vehicle has stopped for more than 10 minutes on the side of a highway or an off-ramp. Fleet managers are notified when this occurs, allowing them to assess the situation and determine whether the vehicle is parked in an unsafe area or is simply stopped in traffic. Managers can then request that the driver move the vehicle if needed. This helps managers ensure the vehicle and driver are safe and not at risk of causing a costly collision.

With our recently released Weather Hazard Alerts, fleet managers can receive real-time alerts based on geospatial and National Oceanic and Atmospheric Administration (NOAA) weather data to help facilitate faster decision making for safer and more efficient driving. Alerts can be configured to cover instances of snow, thunderstorms, floods, tornados, and more. Considering that one in five crashes on average are weather related, these alerts have the potential to make a significant impact on road safety.

Our ever-expanding catalog of innovative event triggers combined with our MV+AI-powered video provides an expanded view of potential risk that could lead to collisions, lawsuits, or customer complaints. By leveraging multiple technologies assisted by professional human review, only Lytx can detect the risk that matters so you can create a safer, more efficient, company.

Click here to learn more about Lytx, or contact us to get started.

* Lytx MV+AI technology is a driver aid only. Drivers should never wait for a warning before taking measures to avoid an accident. Lytx MV+AI distraction detection and alerting technology does not collect, store, or use any biometric identifiers or biometric information (e.g., scans of facial geometry) to detect distracted driving behaviors. See

Essentially, MV+AI works as the eyes and brain of a dash cam. Machine vision sees and identifies driving behaviors. Artificial intelligence uses computer algorithms to interpret what machine vision sees.