PZU, the largest and oldest insurance company in Poland has been our client for three years. The car insurance provided by traditional insurers is increasingly becoming similar to direct companies and it is difficult to prove the value of higher prices.

The whole category of traditional insurers builds the value on the promise of highest client service standards and fair compensations. Direct insurance companies started the price competition providing worse service, however comparative advertising is prohibited in Poland and this fact can’t be communicated. Moreover, the traditional insurance companies are perceived old fashion and expensive. PZU as a category leader is responsible for growth perspectives and condition of the whole insurance market in Poland.

We defined our challenge as “changing the status quo”. We wanted to create not only a solution to the category commoditization, but also an opportunity to provide clearly visible additional value to our client’s car insurance product.

The project was our initiative, so we started with no budget. Until the working alpha prototype three people were involved: Strategy Director, Head of Mobile and Head of Business Innovation. The cost was 15 000 Euro.

To show the difference, we decided to make a difference. If the customer need is “I want to feel safe”, we wanted to make people REALLY feel the very special care.

The direction we took was a result of observation how difficult is the time, when something is going wrong with a car. We concentrated on reducing the time needed to resolve the issue and removing every second of uncertainty possible. Sometimes, even don’t let the anxiety arise in the first place.

The technology we used isn’t new. The idea of a “connected car” was developed in early 90s in Japan, and since then the communication with car’s computer has been standardised and become accessible in all cars produced all over the world. Cheaper cloud computing services and popularity of data mining techniques allowed us to create new value from existing technology and data.

Our device is a simple plug-and-forget piece of technology, which may be used by anyone and does not require any technical skills. Users connect it to a standard car diagnostics connector located in an accessible place in every car produced after 1996.

Then all data from sensors present in a car is collected by the device and sent over mobile network to the computation cloud. Data is processed and analysed real time, giving us time to react, if we notice anything alarming.

When a failure is identified, we can call our customer and explain the danger, immediately send roadside assistance with necessary parts and inform them on the estimated time of repair. The system allows us to react just before or immediately after something wrong happens with the car.

By adding data mining to the standard insurance company components, such as roadside assistance and call centre, we are redefining the car insurance product.

Our goal is to make PZU Drive system an integrated part of car insurance product. It is a self-learning and scalable system, which will be performing better with every new car connected.

To ensure customer privacy, we will NOT use any location tracking module and we will not be able to detect traffic violations. Our modules CAN’T control anything in the vehicle and all data is anonymised to make sure they can’t be used against our customer. These restrictions are critical to the potential of wide adoption of this technology.

Thanks to this solution, PZU may provide the best car insurance and redefine the role of the insurer – from a company processing premiums and claims to a company taking care and preventing bad things from happening.

At the moment we are preparing a test on 100 cars from PZU corporate fleet. Our current goal is to calibrate statistic procedures and prepare them for commercial launch. At this test stage, we will be able to react to every standard fault arising in any car connected to the system. Increasing user base will allow us to evaluate patterns of normal car performance, detect abnormalities and provide more advanced failure detection.