{"id":5410,"date":"2020-04-26T16:52:52","date_gmt":"2020-04-26T14:52:52","guid":{"rendered":"https:\/\/advancedfleetmanagementconsulting.com\/eng\/?p=5410"},"modified":"2020-04-26T16:52:52","modified_gmt":"2020-04-26T14:52:52","slug":"lytxs","status":"publish","type":"post","link":"https:\/\/advancedfleetmanagementconsulting.com\/eng\/2020\/04\/26\/lytxs\/","title":{"rendered":"Lytx\u2019s MV+AI technology explained \u2013 how the DriveCam detects risks"},"content":{"rendered":"
The machine vision and artificial intelligence technology behind Lytx \u201ctriggers\u201d<\/span><\/strong><\/h5>\n

The innovative DriveCam<\/a>\u00ae<\/sup> helps improve fleet safety<\/a> by watching for risky driving behaviors on the road and in the vehicle. The device\u2019s advanced machine vision<\/a> (MV) and artificial intelligence (AI) technology analyzes driver behavior and nearby vehicles to determine how a driver is performing relative to their surroundings. Specific driver behaviors such as following too closely, failing to obey a stop sign, or texting while driving, trigger the DriveCam to flag and \u201ctrigger\u201d the event.<\/span><\/p>\n

What are the benefits of MV+AI technology?<\/span><\/strong><\/h5>\n

Lytx\u2019s MV+AI technology helps fleets by providing an expanded view of risk. The insights from MV+AI can reveal if a driver needs to work on reducing a certain behavior or if certain risky behaviors are appearing among all drivers. If this is the case, an organization-wide initiative to improve safety may need to be implemented.<\/span><\/p>\n

How the MV+AI technology learns<\/span><\/strong><\/h5>\n

Lytx\u2019s MV+AI technology uses information from several sensors at once to identify risky situations and driving behaviors. Machine vision can see and recognize objects and behavior by analyzing images and video data. AI interprets and learns from those images and video data to determine the likelihood that a particular event or behavior occurred. The combination of information from video and other sensors in and around the vehicle, called video telematics<\/a>, provides information to help the artificial intelligence learn\u2014similar to the way our brains rely on information from each or our senses to understand what is happening around us.<\/span><\/p>\n

Not just any data can be used to train the AI to identify and categorize risky driving behaviors. The data must come from a large database. It also must be validated. The most accurate validation comes from human reviewers. Here\u2019s why:<\/span><\/p>\n

Human experts help surface important nuances that help the system learn faster and interpret more complex scenarios. For example, AI technology might detect a lane departure and trigger a video event clip. We know that lane departure is highly correlated with distracted driving such as texting. \u00a0However, a human reviewer might notice that this swerve was due to a construction area forcing all traffic to make a sudden lane change. This type of nuanced data is important in helping to refine, learn, anticipate what might happen next and become more sophisticated along the way.\u00a0\u00a0Lytx events have been reviewed by humans for over 10 years.\u00a0<\/span><\/p>\n

How do MV+AI triggers work?<\/span><\/strong><\/h5>\n

Machine Vision recognizes an object through the DriveCam event recorder\u2019s lens, and AI uses that information, along with other data like speed, GPS to determine if that combination of information is risky. You can think of machine vision as the eyes, and AI as the brain.<\/span><\/p>\n

For example, when a vehicle approaches a stop sign, machine-vision technology scans the sign and recognizes it. The system also uses speed data to detect whether the driver came to a complete stop. If the accelerometer detects that the vehicle didn’t come to a complete stop the DriveCam system triggers a video clip.<\/span><\/p>\n

Here\u2019s a list what the DriveCam can currently detect on the road using Lytx\u2019s MV+AI :<\/span><\/p>\n

Rolling stop:<\/strong><\/span><\/p>\n