ITC DIA Europe

Sentiance & Autoliv: Mobile crash detection that not only manages risk, but also saves more lives

Written by Roger Peverelli and Reggy de Feniks - Founders The DIA Community on Jan 21, 2021

Every minute counts when it comes to saving lives. Being notified in real time about an accident can make a real difference for the people involved in the accident, especially at places where it takes longer to get to the scene of the accident. With crash detection everybody knows exactly when and how severe an accident has been. Sentiance and Autoliv have a shared vision to build the world’s best mobile crash solution that can automatically detect crashes, enable high confidence alerts for first responders, and estimate the extent of damages and injuries as I leave and sentence.

Crash detection is becoming critical for insurance companies, especially mobile crash detection with high accuracy reports and real-time notification. Sentiance created algorithms and models with the latest machine learning technology to manage the technical challenges when developing crash detection solutions. Autoliv’s Safety Score uses Sentiance’s motion intelligence SDK to monitor the movement of a driver’s mobile device. With this app, insurers can actually help their customer at their hour of most need.

Sentiance and Autoliv’s crash detection is grounded in science, driven by data and the latest machine learning technology. To an insurer, an effective crash alert system is huge and potentially game changing. Firstly, it offers significant customer experience benefits because it places control of the claim in the hands of the insurer and removes any inconvenience from the customer in terms of reporting the claims. You can help out the customer and take control to claim the second it happens, which offers huge values to the customer. Secondly, it drives commercial benefits. As an insurer, you’ll get real time notifications and you’ll lower your costs by automating the claims settlement process with reliable data. You can understand what causes crashes through our contextual data, and you can use these insights to prevent future crashes.

How it works
Autoliv’s Safety Score uses Sentiance’s motion intelligence SDK to monitor the movement of a driver’s mobile device and gain certain insights into their driving events. The events are then further processed and validated through Autoliv safety knowledge, providing the user with a personalized 3-digit safe driver score. Drivers with higher scores exhibit more safe-driving behaviors over an array of data points including turning, acceleration, braking, speed, and distractions.

Crash detection
Crashes are hard things to study. Sentiance and Autoliv identified three levels of crashes and developed critical intelligence that is needed in a crash detection service. The three levels are:

  • High severity: we typically see an airbag being deployed and serious injuries as a result of the accident. Sentiance can tell through their algorithm, to an accuracy of about 90 percent, if an accident has happened or.
  • Medium severity: an airbag could be deployed, but it doesn’t have to be the case. Cost of fixing and repairing the vehicle adds up to fifteen hundred US dollars or more.
  • Low severity

In crash centers, real car crash tests are where smartphones are positioned throughout the vehicle. Through that, Sentiance gets validated and ground truth data that they can use to train their algorithms, but also use to understand the context and other factors influencing when a crash happens in that spectrum. Separate events that happen on the phone, become crucial in solving crash detection. Only through proper testing, validation and scoring this can be done. Understanding that pattern, reaching out and interpreting that to give you the right data at the right time is what Autoliv and Sentiance does. This is true science behind crash detection.

The application of the crash solution will differ depending on a client’s use case. Sentiance and Autoliv are building three crash detection models to address low, medium and high severity crashes. They train their crash detection models with a variety of data sets and we use real crash data. They also use millions of trips that we’ve collected over time to recognize typical false positive scenarios. This results in a highly accurate crash detection solution and models that are constantly improving.

Contextualisation platform
Detecting a crash at the time of impact is critical, but insurers are also interested in understanding what happened before, during and after the crash occurred. This solution can identify the context of the trip. Is it a commute to work or is it a leisure trip? What was the drivers behaviour before the crash? Were they speeding or were they distracted by using their phone? With the contextualisation platform the insurer gets all these relevant insights in the crash report. With these insights, insurers can get an in-depth overview of the crash, understand what caused it, and speed up the claims settlement process.

Who are Sentiance and Autoliv?
Former DIAmond Award Winner Sentiance is a team composed of the best and brightest minds across multiple disciplines, such as data science and engineering, determined to create the next generation contextual platform at global scale. Toon Vanparys, CEO of Sentiance came on board of the company as early seed investor and has acted as CEO as of October 2014.

Autoliv is a worldwide leader in automotive safety systems, with sales to all major car manufacturers. Autolivs more than 65,000 associates in 27 countries are passionate about the companies vision of Saving More Lives. Autoliv has been responsible for major technological breakthroughs within automotive safety for more than 65 years. Since 2018, Mikael Bratt holds the position of President and CEO.

Website: &

Check out their live demo here:

Amsterdam 12-13 June


00 Days
00 Hours
00 Minutes
00 Seconds
Have a look!
ITC DIA Europe - Amsterdam 2024 - Register here! Purchase now!