The entire automotive industry is heavily invested in the race towards its holy grail, the self-driving car, with several OEMs targeting 2020 as the year in which we will finally start seeing autonomous vehicles cruising our streets.
While many experts believe these cars will be too expensive for the average consumer to afford, it’s important to understand that autonomous vehicle technology will continue to evolve rapidly, bringing costs down for the average consumer.
Today, we are already seeing how new technology is making driving safer without breaking the strict budget constraints of consumers.
Now more than ever, advanced algorithms and machine learning techniques are opening the door for the introduction of leading-edge safety and convenience features.
Today's average new car is already equipped with a long and diverse list of sensors: front facing cameras for lane keeping functionality, radars for blind spot detections, rear view cameras for parking assist, and the list goes on and on.
We at Telenav have been working hard on fusing these sensors' information with the more traditional car sensors (e.g. GNSS positions, gyro, barometer, speed etc.) and the standard navigation map in order to reach Lane Level Position accuracy.
In other words, we can leverage the existing equipment in an average car to accurately determine the car's current lane.
To help explain the potential behind this technology, please take a look at the following YouTube video (warning: not easy to watch).
The driver in the video was trying to exit the highway from the right-most lane. It is also clear the person was not driving carefully, switching lanes way too quickly. Finally, during the last switch, the big yellow truck blocked the sight of the driver, who was probably assuming the right lane was progressing at the same speed.
But this accident was largely avoidable, and technology that already exists could have prevented it.
As described earlier, Telenav can accurately determine the car's current lane. In this case, the third lane out of three. Additionally, Telenav's personalized route prediction engine allows the system to calculate with high confidence which route the driver must follow regardless if he explicitly keyed in a destination or is driving in a free drive mode.
Different real-time traffic providers can already today deliver traffic information in lane level granularity and these services will just increase in accuracy and coverage once cars can localize themselves more accurately.
Finally, the relative location of the yellow truck would also be "known" through the car's proximity sensors.
Fusing all this information together allows Telenav to offer an advanced navigation experience that would help drivers switching lanes based on their exact lane position and their predicted destination, in a safe manner and would warn them not do so if conditions are unsafe.”
The list of scenarios that could be tackled with this technology is long and includes cases such as: "exit highway", "join highway", "lane end" and "avoid turn only lane" just to name a few.
With the DNA of a mobile company, Telenav is extremely focused on user experience, relying on intensive research iterations with users. Dynamic Lane Level Guidance is a feature that requires such attention to details. Since drivers must perform complicated maneuvers in split seconds, it is crucial that our technology not hijack their attention, but rather help drivers perform the task at hand.
We chose to target the HUD and cluster displays for such tasks. Here is a video of our product from Telenav that demos the exit highway scenario.
To summarize, we are helping OEMs deliver advanced safety features and delightful experiences without increasing the car's cost, by applying advanced localization techniques and by better utilizing the sensors already installed in the average car.
Dynamic lane level guidance will improve the driving experience while keeping us all safer.