In a series of articles about the quiet achievers of the new age of automotive use, here at Motoring Weekly, we have highlighted the work at drive.ai. Now we have another company to be aware of: Nauto.
Nauto was founded in 2015 with a view to getting modern cars and drivers to be safer and they have given themselves a huge target: zero traffic accidents. What sets them apart from the rest of the crowd of engineers, software developers and data scientists is that Nauto recognises that humans will still be involved in controlling a vehicle for decades to come.
That one difference means that they have focused on developing technology to see where humans make mistakes and therefore can help to teach a driver some better habits. In other words, the systems are looking for distracted drivers.
Initially designed for fleet operators, where accident and repair costs are a budgeted and often costly amount, the Nauto Intelligent Driving System uses artificial intelligence, cameras and other technology to understand what the driver is doing, where they are and the prevailing conditions. It collects a wide range of data points and only sends back relevant data to the fleet’s control centre when an incident of distraction or the breaking of a road rule is occurring. That means that if the driver controls the vehicle within the boundaries of the law, then only basic data will be collated.
Some examples of the data collected are:
Geographic location and relative proximity of a vehicle to other vehicles
Hazards, such as pedestrians (whose faces may be blurred), bicycles, weather, and other vehicles
Collision, near-miss, hard braking, acceleration or cornering events
Traffic patterns, speeds, and flow
Lane and road departure
License plate numbers (which may be blurred) and other information such as colour or make of surrounding vehicles
Driver behaviour and actions or inaction
Risks
Number of passengers in the vehicle
Drowsiness
Inattention (e.g., phone use, passenger activity)
Braking
Speed, acceleration, and mileage
Driving time
Nauto device tampering
Vehicle security (e.g., audio listening of broken glass, stolen vehicle)
Not obeying the rules of the road, e.g. texting while driving, ignoring traffic control signals or signs, DUI (i.e. drink-driving)
This list was found in their privacy section of their website and is not the complete list – they can capture enough data to provide immediate notifications to the driver as well as send back the most pertinent data points for further analysis.
Clearly with today’s privacy laws and with the European General Data Protection Regulation (GDPR), Nauto and its customers must be very strict about what is collected and what is distributed, so much of the general data is aggregated and de-identified such that trends can be analysed and general programs developed to help with ongoing education of the drivers.
Nauto is also looking at how the aggregated data can be used by other developers for input into autonomous vehicle databases. This is a great idea because it means that the autonomous system has an extra input in to the decision making brain of the vehicle.
Some of their partners include BMW, Toyota and General Motors who presumably can use the data collected in their own systems – for example, it could be a good input for GM’s Cruise Automation, and even Allianz, the large German insurance company is a partner. This is an interesting connection, because if Nauto achieve their goal of zero collisions, it would mean that the risk of insuring a vehicle would be greatly reduced which in turn reduces premiums and that leads to lower revenues for the insurance companies!
This technology could easily be repurposed by Government departments who are dealing with drivers who have lost their licence through consistent failures to obey road rules and local laws. The technology could be retrofitted to a vehicle and the driver tested to see if they have genuinely learned anything from losing their licence. It sounds a little “big brother-ish” however, if we could all be sure that every driver on the road had a better sense of risk and a higher standard of behaviour, then driving in our towns and cities might be a little more safer.
Logo owned by Nauto.
VinceS2 says
I recently got a Dell for engineering VR application and discovered how extremely good eye tracking is now; here is an example of a similar application to Nauto’s: https://www.tobiipro.com/insight/solutions/professional-performance/
I care a lot about this topic as I am tackling it from the other end; getting people to consistently make better choices, over the long term. I think a 20% reduction in the road toll is eminently doable but, for reasons well grounded in science, won’t be moving to zero as a goal any time soon! But a blend of genuinely changed attitudes and supportive technology mentioned in this article, we should get a lot closer to that end of the spectrum. You can get an idea of how here: http://www.vrdriversim.com.au; whilst I am not updating that website out of IP leakage concerns, the key take-away is that the system is being developed to be economically and practically accessible to the population (and later mandatory), and solidly based on proven science, some of which we will be developing as part of the process.