Mobility plays a vital role in our socio-economic development.
Today, there are over 1.2 billion vehicles on the road, with an estimated 2 billion by 2035. It’s therefore imperative to understand and mitigate mobility risks to make our roads safer.
But with the number of vehicle miles traveled (VMT) per person generally trending upwards, especially with new investments in infrastructure (i.e. the US Bipartisan Infrastructure Deal), and new modes of transportation emerging, mobility risks are constantly evolving.
Drivers traveling more miles per year is leading to increased risk exposure. But what’s making the situation even worse is changing driving behaviors.
Despite a 55% decrease in miles traveled in 2020 due to the pandemic, we saw a 63% increase in collisions per million miles. That led to the highest YoY death rate spike in almost a century.
Several telematics programs have attempted to mitigate mobility risk, but have only been able to scratch the surface. Here’s why:
In order to mitigate mobility risks for a specific region or at a worldwide level, we must adopt a holistic approach and gain visibility into multiple, varied risk factors.
There is a global need for technology platforms that can ingest large troves of data from ubiquitous sources at scale.
Mobility dynamics and risks are constantly evolving with:
But until now, existing technology platforms haven’t been able to adequately understand and mitigate the evolving mobility risk at scale.
If we fail to comprehend and proactively mitigate these risks, our roads will continue to get deadlier.
Collecting large troves of data is one thing, but understanding and processing that data to draw meaningful insights is another. Thanks to recent technological advancements, we finally have the technology today to do it all.
But let’s back up for a minute, and hone in on an important analogy to help us fully grasp the problem at hand, and how we got here in the first place.
We can compare the technological advancements that help us better understand and mitigate mobility risk with data, to advances that help us understand how our own brains function.
Surprisingly, understanding evolving mobility risks is akin to understanding how the human brain works. Here’s how.
A human brain has over 86 billion neurons that interact with one another through complex subsystems and clusters. Understanding any specific cluster isn’t enough to fully comprehend cognition overall. But getting a full view of the brain and human cognition is quite difficult to achieve without access to comprehensive tools.
To understand cognition, neuroscientists study how these neurons communicate with each other from neuron-to-neuron, neuron-to-cluster or network, and, more holistically, networks-to-networks (or the brain overall).
Just as one neuron cluster doesn’t help us understand the entire brain, one cluster of vehicle or driving data doesn’t give us a complete view of the diverse multitude of mobility risks.
Sure, it’s possible for one small pool of vehicle or behavioral data to speak to a specific segment, but it can’t show how that segment’s risk relates to or is impacted by the risk of several other segments.
To understand the multi-faceted nature of mobility risk, we need to analyze billions of miles of mobility data coming from diverse segments and sources.
If we delve a bit deeper, we realize that the study of neuroscience and understanding mobility risk also share a similar historical evolution.
Phrenology and the first wave of telematics
In the early 19th century, Franz Joseph Gall, a German neuroanatomist, introduced the science of phrenology, which suggested that a person’s character or personality traits can be determined by reading the bumps on the skull.
Although it’s now considered pseudoscience, at the time phrenology was introduced, people used to detect signs of latent delinquency in children or elect politicians based on the shape of their skulls.
By the same token, the first wave of hardware-based telematics solutions attempted to understand mobility risks. However, it barely scratched the surface because the hardware was focused on the physical properties and movements of the vehicle, rather than human driving behaviors.
Due to the inherent complexity of these hardware solutions and their consequential limited adoption, they failed to collect reliable data on key risk factors. This reliable data, of course, was essential to truly understanding and mitigating mobility risk at scale.
Following phrenology was electrical brain stimulation (EBS). By implanting electrodes in the brain through surgery, neuroscientists tried to collect localized data to understand brain function.
However, EBS wasn’t an effective way to fully understand its function. First, very few EBS studies were carried out due to the invasive nature of the process. Second, electrodes locally placed in one specific area of the brain could only provide data on neurons in the immediate vicinity of the electrodes. It was therefore difficult to measure the impact and relationship of one segment of the brain with other segments.
Comparably, what came next in the telematics’ history were first-generation smartphone telematics programs, which provide data on a specific segment of drivers on the road. But that data is naturally skewed, since it’s collected from homogeneous sources and use cases.
For example, first-gen programs might measure data on tens of thousands of drivers in a rural location, which wouldn’t help us understand driving behavior with the same flexibility and resistance to biases as would programs with data on hundreds of millions of drivers globally.
In the late 20th century, the advent of magnetic resonance imaging, or MRI, proved to be a game-changer in the field of neuroscience.
Researchers could finally get a clearer, more holistic understanding of brain structure, and later with Functional MRI (fMRI), even the brain activity and blood flow. By looking at the Blood Oxygen Level Dependent (BOLD) contrast, they could detect higher neural activity in the areas of the brain that were in use as the oxygenated blood flow to that specific region increased.
With the help of fMRI, neuroscientists can now clearly understand which actions and emotions trigger activity in specific areas of the brain, as well as the role different parts of the brain play in cognitive function.
Just as MRI changed the way we understand and diagnose medical issues by getting better visibility into pre-existing and potential health risks, Zendrive is transforming the ways in which we identify and understand mobility risk.
We’ve built a platform that provides a complete view of mobility risk by ingesting and analyzing diverse data from multiple sources. We call it Mobility Risk Intelligence, or MRI.
Unlike other platforms that provide limited insights on small subsets of vehicles and drivers, Zendrive’s Mobility Risk Intelligence (MRI) platform gleans data from tens of millions of drivers across the globe to provide comprehensive insights on multiple pools of drivers and their impact on mobility risk.
By amassing one of the largest mobility datasets in the world, the MRI platform is continuously identifying new risk factors, making it easier for its users to understand and mitigate the ever-evolving mobility risks.
At Zendrive, our mission is to make roads safer through data and analytics.
By utilizing the most ubiquitous device in the world — the smartphone — we ensure our users’ safety by turning raw sensor data into risk signals.
We understand that with the introduction of new mobility options and modes of transportation, our users may change their trip mode more than once during the day. It’s therefore important to ensure their safety regardless of their mode of transportation.
Our MRI platform is powered by patented AI and machine learning algorithms that are trained by over 200 billion miles of data collected from millions of drivers across the globe.
Today, Zendrive’s MRI platform is helping tens of millions of people improve their driving behavior, get fairer insurance, seek immediate automated help in case of a collision, and keep an eye on their loved ones’ driving behaviors.
The algorithms powering Zendrive’s MRI platform are trained by the richest driving dataset in the world. What’s more, the quality and quantity of data is ever-increasing, helping the MRI platform build highly predictive and contextual risk models.
Let’s take a closer look at the factors that set us apart on the market.
Since our inception, we believe that the future lies in ecosystem services. We partner with the leading Automotive OEMs, consumer apps, mobile network operators, fleets operators, and insurers to get mobility data from a wide variety of sources. This helps us ensure that the data is diverse at its source and not skewed toward a specific geography, vehicle, driver, or use case type.
Our partnerships with consumer apps provide us with access to fresh data representing all types of drivers at unprecedented volumes.
With hundreds of millions of users across the globe, fresh data gathered and analyzed on an ongoing basis helps us frequently update our models and understand specific route and location risks.
Unlike telematics providers who mainly offer usage-based insurance (UBI) programs that have historically led to limited adoption and high churn, our MRI platform uses behavioral data captured for long periods of time, allowing us to take seasonality into account.
This helps us understand how the same user drives during different times of the year under varying weather conditions. The MRI platform takes these subtle differences into account and adjusts its models accordingly.
Our driver distribution is in line with the actual distribution of drivers. With its highly diverse user base, the MRI platform takes regional behavioral differences into account.
As Zendrive is providing its technology to millions of drivers across the globe, we have been able to test our technology rigorously on different device types. Having performed tests on over 6,000 devices with hundreds of SDK versions deployed, we developed a strong understanding of sensory disparities between different operating systems and device types. The MRI platform takes these differences into account and makes adjustments accordingly.
The MRI platform is trained by the largest smartphone-centric collision dataset in the world.
With hundreds of thousands of collision and non-collision events analyzed, the MRI platform detects collisions with a remarkable accuracy of up to 95%.
The MRI platform is therefore able to build highly predictive risk models with actual collision data tied to real-time behavioral data.
With its robust risk models, the MRI platform is improving the way people drive by providing timely, accurate feedback on driving behavior through driver and vehicle scores. These contextualized mobility insights - which can be used in engaging insurance reward programs - lead drivers to make smarter, more informed decisions.
According to Milliman — the leading actuarial consulting firm — Zendrive’s score is 6X more predictive than the industry’s leading scoring models. In a study of fleet drivers, the MRI platform reduced the collision risk by 49% by providing timely feedback to the bottom quartile of drivers.
The MRI platform is helping insurers across the globe build advanced frequency and severity models by combining different types of data generated from different sources for different regions.
Insurers can leverage the platform to price and segment risk precisely, and make better decisions regarding risk coverage and reserves, helping them ensure long-term profitability.
Zendrive partners with system integrators and solutions providers to ensure a seamless flow of mobility data across different platforms.
The MRI platform is helping millions of people get personalized insurance quotes based on their driving behavior.
The MRI platform is helping its partners provide contextualized help to their users in the moments of truth before a road accident, within a timeframe of 20 seconds.
Insurers can also reduce fraud and claims severity and shorten the claims cycle by getting critical claims data and insights within seconds of a collision.
The platform allows friends and family members to keep an eye on their loved ones’ driving behavior. In case of a collision, the platform notifies the family circle. It also allows family members to create geofences, speeding, and phone usage alerts.
With its robust data, the MRI platform also equips city governments with contextualized insights to make smarter decisions around mobility. City officials can also mitigate mobility risk on an ongoing basis by providing safer driving programs to their citizens.
The MRI platform is providing fleet managers with visibility into their fleet’s behavioral and vehicular risk. It can also develop risk models specifically for trucking.
The MRI platform is enabling research for route safety studies, which will help automotive OEMs, commercial fleets, and insurers globally. This research will play a key role in enhancing the predictiveness of future risk models.
The MRI platform will enable a smoother, safer, and faster transition to an autonomous vehicle era. Zendrive will continue to help its partners leverage its proprietary capabilities to ingest and analyze device-agnostic data in near real-time.
With the evolving landscape of mobility, the nature of mobility risks will continue to change. While the world prepares for another mobility revolution, Zendrive’s MRI platform is constantly refining itself to understand and mitigate new mobility risks by providing preemptive risk signals and insights to its partners across the globe.