Driver Safety

December 8, 2020December 8, 2020

Insights from Telematics Projects Across the Globe

Allied Market Research recently published data pointing at a 34% CAGR in the usage-based insurance (UBI) market between 2016 and 2020. Such a rapid expansion would be nearly impossible without some form of collaboration between automobile manufacturers, insurers, and telematics technology platforms.

Understanding how the three have collaborated across the globe in various pilot projects and initiatives can highlight critical insights, frameworks, and guidelines for other insurers contemplating the launch of a UBI product. While there have been several prominent collaborations across the UBI value-chain in different segments, here are some of the notable collaborative projects which have yield positive results for the involved entities:

1. Using Driver Behavior as a Metric for Market Segmentation

The Floow UK, a telematics developer, produced research showing the inherent demand among the insurers for driver behavior data. Over 32% of insurers were seeking access to such specific datasets to understand the underlying risk profiles of drivers and tailor the insurance profiles in-line with these risks.

Driver behavior attributes like hard braking, sharp turns, relative speeding, and acceleration are considered key in gaining insights. More comprehensive platforms like Kruzr can gauge other impacting attributes like fatigue and distracted driving as well.

Conventionally, actuarial models have been the key mechanism to gauge the risk attributes for insurance applicants. Checking whether the risk is accurately priced or not is an altogether different discussion. Still, when the risks are not accurately priced, it can drive the consumer away and result in the loss of potential revenues.

Given the fall in profits from the non-investing side of the business and highly concentrated revenues in the market, it is apparent that most auto insurers cannot afford to let potential revenues slip away with mispriced risks. Hence, the primary level of adoption is coming to segment the market beyond the actuarial models. Instead of using just the demographic data, geographic data, car’s details, and credit scores, insurers are seeking telematics data that helps them understand the risks associated directly with the driver’s behavioral attributes.

Companies like Root Insurance have shifted their entire insurance risk pricing model to gauging driver behavior. The company offers insurance premium estimates in under a minute and offers annual savings of up to $900. Other than that, the company has worked extensively on ensuring the claiming process is simple and paperless.

2. Exploring New Revenue Streams with Data Marketplaces

Data Marketplaces are already existing in capital markets services that use alternative data and marketing communications services where email databases are frequently used for sending out mass emails.

McKinsey forecasted $1.5 trillion in incremental revenue pools for the entire automotive industry with the help of connected cars. Telematics and the platforms using them will be a major component in this revenue growth.

Automobile manufacturers seek additional revenues by adding data-capturing capabilities to their vehicles right from the point of manufacturing. Ford Commercial Solutions, a division of the Ford Motor Company, recently announced its data services launch. The company has started installing modems into its commercial vehicles. Data on vehicle health, driver behavior, GPS, mileage, and fuel usage is collected by the modems and uploaded directly on a cloud platform, engineered and maintained by Ford.

The same data is then made accessible as an on-demand service to telematics companies and fleet owners who either use it for direct analytics or enrich their product offerings.

The insurers might be on the buying side of data if they are not already running UBI programs independent of modems and black-boxes pre-installed by automobile manufacturers. Insurers, who possess the consent from consumers, might monetize that data by making it available to advertisers.

Kruzr collects a wide range of data-points such as hard braking, acceleration, speeding, sharp turns, drowsy behavior, and fatigue. This data is then augmented with GPS, traffic, and road conditions data. The insurers and automobile manufacturers who want to use this data can serve as an additional line of revenue, with advertisers and other industry incumbents acting as the key buyers.

3. Telematics for Product Development and Features Outside Insurance

As per a Frost & Sullivan report, the number of connected cars with embedded telematics has been projected to grow by nearly 3x between 2018 and 2025. While the apparent application of this data would be in the form of usage-based insurance products, the same data is being leveraged for other use cases.

The telematics data available with Hyundai Ioniq and Kia Niro provides drivers with real-time information on charging information, nearby stations, and chargers’ availability. Honda has taken this a step ahead with its SmartCharge Beta Program. The program uses telematics data, and the data pulled from the city’s grid to direct the drivers for charging their cars when the power demand is low in the city. This helps the city’s power grid maintain a consistent power supply instead of working with sub-optimal calibrations made for fluctuating demand. It helps the drivers get quick access to already available power, which other consumers cannot use.

Kruzr SDK has been optimized to provide white-label solutions that let the automobile brand take the center-stage. The data availability, navigation, weather data, and traffic congestion insights are already aggregated to calculate real-time risk. Adding more geospatial data for expanding the use-cases is an easy integration possible with the platform.

4. Creating a Trust-Based System and Accelerated Claims Processing

Ping An, one of the largest automobile insurance companies globally, has an internal team focusing on the applications of AI and how it can be used in conjunction with telematics data to provide better auto insurance products.

The company has recently deployed two key offerings. The first one focuses entirely on gauging the driver’s behavioral attributes. The driver is given an individual scoring in terms of trust quota, using big data and AI. This trust quota is used in streamlining self-help insurance claims.

The second offering focuses on expediting the self-help process. Right after an accident, the customer is asked to click photos of the car, enter the amount of the repairs and the damage incurred. Then, using the trust-quota to weigh the sincerity of the claim along with deep learning and visual computing, the Ping An app determines the extent of covered damages. The system closes the process in 3 minutes between clicking photos and deciding on the claim application.

In Conclusion

Telematics should be looked at as a strategic initiative that can expand into new markets, provide better services, and create more optimized insurance products. You can virtually use the telematics data for delivering value in any direction of the value-chain, with a comprehensive platform like Kruzr.

August 18, 2020August 19, 2020

Paradox: Cellphone Distraction Prevented by Cellphone

Does active driving guidance which functions in real-time on smartphones serve more as a distraction and less as a safety tool? It is a popular opinion that interaction with cellphone of any kind implies danger.

The reasoning to back this is based on the fact that time and attention necessary to interact with the driver safety guidance system can be distracting.

That argument makes intuitive sense. And it was one of the key premises used in a case filed against a leading ride hailing app, which is yet to receive a decisive judgement. The research published by a team from Columbia University showed that using the ride-hailing app resulted in increased accidents during the year-long analysis.

While that premise may or may not hold true, extending the same logic to active driver guidance system that operates in real-time falls halfway. Active driver guidance system like the one deployed by Kruzr operate on an audio interface that alerts the driver instead of making her/him leave the steering or look away from the road.

Real-time alerts are delivered throught voice to the driver to prepare him/her for the upcoming danger. These alerts range from speeding, to accident hot-spots.

In addition, it has the capability to allow only emergency messages and calls to get through with an advance notice and advice to the driver to halt for answering the incoming call.

On the other side of the argument are the ride hailing apps. The research inferred that the app was motivating drivers to make more trips and to pick up individuals from congested areas. Whereas, Kruzr is designed to motivate users to stay safe and avoid physical interaction with the mobile device in congested areas and at high speeds.

The Value of Real-Time and Data-Driven Driver Training for Insurers

Driver training is generally looked at as a risk-aversion mechanism. By reducing the probability of a crash, it allows insurers and reinsurers to mitigate the risk of paying for a claim.

Kruzr’s platform has been designed to deliver value beyond that. By alerting the driver and nudging her/him for a change, a large portion of the risks can be mitigated. Along with that:

1. Personalized Risk Assessment While the driver is given the necessary alerts, the data collected can help the auto insurer understand the driver’s risk profile at a deeper level. Several data points that get collected over time using driver’s own behaviour on the road can help the auto insurer offer more risk- sensitive policy premiums that motivate the driver to be safer and help the insurer avoid claim upticks.
2. Competitive Offerings Paired with the telematics system on the platform insurers can provide more than just insights on driver behavior. They can actively help drivers identify risk while on road and prepare them for life-threatening scenarios which would go unseen without active safety guidance.
3. Systemic Risk Mitigation Each driver has to deal with the systemic risks such as traffic & congestion, accident-prone zones, and weather conditions. Most conventional driver training systems would not be able to take the driver through real-life scenarios in order to prepare her/him for them. Kruzr actively alerts the driver about forthcoming risks and can also help the driver navigate through one, if that is the only option available.

In conclusion, real-time alerts from an active safety guidance system are as much as a source of distraction, as sign boards on the side of the road, alerting the driver of accident hot-spots ahead, speed-limits, directions etc.

August 10, 2020August 11, 2020

Driver Coaching: An Essential for Driving in the 21st Century

Auto Insurance is a business of optimizing and controlling variables. The more control you have over certain variables, the more risk you will be able to mitigate. This would eventually lead to more competitively priced policies, higher market-shares, and eventual industry-leadership.

Focusing on Driver Coaching is one of the many measures available to insurers, reinsurers, and the industry, to mitigate risks at the driver-level.

The Need for Driver Coaching

If one analyses just the fundamentals, there should not be a need for driver coaching. The government issues licenses to state whether a driver has the necessary skills for driving or not. When used along with background-checks for traffic offences, companies should be able to filter high risk drivers.

However, such measures are ineffective because accidents are not caused by how a driver reacts to an everyday situation, but by how the driver reacts to a specific unforeseen situation. The numbers make a clear case for the crashes and injuries inflected by driver behavior as a cause:

Alcohol, speeding, and distraction are directly responsible for over 57% of accidents caused every year in the United States.
While the number of such crashes might not sound alarming, on an aggregate basis, crashes account for injuring over 4.4 million people every year.

As an auto insurer, these numbers reflect business-risk for you. Increasing quantum of such accidents can result in claims going through the roof for physical, property, and legal damages. The number of crashes is big enough for most auto insurers to worry about. But looking at the probable causal factors also shows one more insight – that these accidents are related to inadequate driver training.

Assuming a driver, who drives professionally or for personal reasons, has undergone effective driver training, she/he should not get into accidents caused by behavioral issues like driving under the influence of alcohol, speeding beyond the safe limits or distracted driving. The question remains – since there is enough driver testing done before issuing a license as well as a good number of drivers training available, why are the crash figures so big?

Deficiencies in the Existing Driver Coaching Methods

The answer to that last question lies in understanding the existing driver training models. Most drivers undergo training because they have been asked to do so by their employer or someone of equal authority. Or, they are about to appear for the drivers’ license test and are hence going through the training with a professional instructor.

The key issue lies in the way such driver coaching methods are structured:

Frequency and Tenure of Coaching: Most of such driver-coaching programs are conducted by independent contractors or companies offering similar services. However, these are one-time training programs. Most of the people who attend such workshops or programs do so to attain their license and once they have it, they are not really concerned about the concepts they were taught. Convenience takes over safety and driving safely becomes a subjective matter.

There are no checks and balances to rectify the drivers not following the guidelines of driver coaching modules. The authorities act after an incident or breaching of the driving rules have taken place, not entirely before that.

Most Programs Use Boilerplate Information: While this information might be in line with what is prescribed in the government safety standards, they do not consider each drivers’ reflexes and patterns in driving attributes. Such workshops can tell you generally what a driver should be doing, not what you specifically should be doing. Most of the crashes are caused by a conjunction of factors out of driver’s control and the ones directly emerging from her/his driving attributes. Generally, the programs using boilerplate information focus only on the former and the risk posed by the latter remains unmanaged.

Solution: Data-Driven, Real-Time Active Guidance

Having an integrated platform that can deliver multifaceted driver training is the key to solving the problem of increasing crashes. Here is what an ideal platform of this type would look like:

High Level Data: It should alert drivers about congested areas that might be prone to high probabilities of accident. An off-the-shelf telematics system can only show in retrospect whether the area was prone to accidents. An effective system that focuses on driver-training in real-time will help the driver actively avoid such situations in the first place.

Nudges to Accommodate Comprehensive Data Points: Speed increases the probability of a crash since it does not give you enough time to react in case something unexpected happens. Most driver training programs will show the driver what is the ideal speed one should drive at depending on the road and traffic conditions.

A lot of drivers who get into a crash due to high-speed might not understand immediately when they have breached the driving speed limit. A system that can take inputs from traffic, weather conditions, road quality, and speed limits will be able to help the driver develop an intuitive sense of when she/he has breached a safety limit and nudge her/him to bring it under control.

AI-Based Predictive Analysis in Real-Time: Machine learning algorithms that take granular data from past driving patterns, trends indicating slowing reflexes, and driving variation analysis in real-time to predict drowsiness can dramatically reduce crashes and keep the driver safe.

Driver Training is essential to reinforce safety standards. But using a textbook program to deliver this information in the absence of any predictive checks & balances would not yield the necessary changes.

By using Kruzr’s integrated platform, you can help your insured customers develop safe driving habits that are actively engineered in line with their own driving behavior, while still ensuring they accumulate an intuitive understanding of the standard rules & regulations.