One of the key questions we are asked is how do we test our product?
Everything we do in our product - the fonts, colors, time to show events, and of course what events to show are meticulously thought through with the idea of making the driver safer. Here's a couple examples of how we make our product.
How do we decide when to show an alert?
Much or our work requires deciding where to show an alert. We coorlealate accidents, road geometries and dozens of other factors to determine when to show information. This information is gathered from public data, our partners and cities private data, and hired folks who travel the city looking for critical situations.
We also take into account the driver, modeling speeds and behaviors on when to show information. For instance, our research from this model shows that on this curve, if you mis-judge the available friction by just 3%, the deviations from the path can be as much as 12m - causing a spin out. This allows us to pinpoint exactly when to show information.
How do we decide what to show?
Using sophisticated vehicle simulation labs, we are able to test everything from time to show information to what is shown.
In the example below, drivers were shown a no win scenario - where nearly everyone was distracted and missed a pedestrian crossing. By showing the right information and the right time, drivers with Signal focused on the road and were better at avoiding the accidents.
We take safety seriously. Each test has its limits, and we constantly get feedback from users and use our product ourselves. Drop us a line and let us know if you ever want to be a tester!