How Fleets Are Utilizing Artificial Intelligence (AI) Today

11 octobre 2019

Technology in trucking is rapidly transforming every day. Every year, the number of devices on each truck and the safety systems that track them are growing. As the amount of technology on the road and inside the back office expands, innovations are being put in place to bridge the gap. The latest revolution in fleet safety is Artificial Intelligence (AI). Using platforms like the Idelic Safety Suite, fleets are integrating all their systems into one location, predicting at-risk drivers, and preventing accidents. Here’s how it all works:

Systems Integrated

Driver performance starts with data. For AI to predict an unsafe driver, it must first analyze all the data collected through integrations. Systems like the Safety Suite start by integrating all third-party safety systems into one place. Just a few of these include:

Telematics
Onboard Video Systems
HR Systems
Accident Records
Learning / Training Management
CSA Violations
Background Checks
Drug / Alcohol Tests
Asset Management
Sleep Apnea

Once all of this data is captured and integrated into one centralized location, Machine Learning (ML) steps in.

Predict Accidents with Machine Learning

ML algorithms predict at-risk drivers with the safety data integrated into your fleet’s AI system. While AI, ML, and predictive analytics may seem confusing, integration and automation make Safety Managers’ lives very simple.

What is Machine Learning (ML)?

Machine Learning originates from the idea that you can give machines access to data and let them learn for themselves. ML can absorb massive amounts of data to build models around patterns of behaviors. Systems with ML can automatically learn patterns without being explicitly programmed. As the ML model evolves, it can then be fed new and live data to predict those same outcomes for what has not yet happened.

The real magic behind this technology is assigning weights to the events and automatically determining which of those are significant predictors. ML takes these weights and safety events and runs thousands of simulations to match the appropriate weight to each event. Trying this by hand is impossible and incredibly inefficient, which speaks to the true power of ML.

Algorithms Based on Past Data

In feeding pre-accident data into a model, data is run through a cycle to identify which drivers it would have predicted. These are compared to the actual outcomes to depict accuracy. Then, the model adjusts the weights of the inputs and reruns it. If the accuracy improved, it knows it’s on the right track. If it deviates, it starts to see why. It then slightly tweaks the weights again and runs another cycle. The model goes through this process thousands of times, ‘learning’ as it goes until it finally arrives on the optimal weights.

Driver Watch List

Once the integrated data runs through the ML models, the system determines which drivers are at-risk for an accident and assigns them a risk score. The Idelic Safety Suite’s AI-generated Watch List is easy-to-read and includes details on the safety events which flagged the driver, each driver’s risk level, and their exact score. Best of all, these processes occur behind-the-scenes, so you can leave the technical jargon aside.

Performance Improvement & Accident Prevention

The actions taken to prevent accidents are equally as important as the ML watch lists that are predicting at-risk drivers. The Driver Watch List can be used not only to view your most at-risk drivers but also to correct behavior before it’s too late.

Assigning targeted performance improvement plans based on specific driver events and behaviors is vital to preventing accidents. Fortunately, systems like the Safety Suite allow you to easily manage dates and timelines, assign targeted training and coaching, and carefully track driver improvement.
By identifying drivers who exhibit poor performance and being proactive in remedial correction, top fleets are reducing risk by predicting and preventing accidents.