Artificial intelligence and machine learning have proven to improve the traffic in different cities around the world, both in terms of better traffic management but also in terms of better designed roads and infrastructure. We are looking at less congestion, fewer accidents and lower travel times.

We have now reached a technological paradigm where it has become feasible to use artificial intelligence to solve problems that simply are too difficult with traditional deterministic and statistical models. Today, we can use neural networks to process vast amounts of video data, not just for detection of simple objects but also for analyzing sequences of events. Something that was considered as impossible just a few years ago.

The number of sensors being in use for traffic analysis is growing fast and more and more information has become available to us. We have information from public transits, reporting of traffic crashes and incidents, road camera feeds, onboard diagnostics information from individual vehicles, navigation data, localized weather data, booking data from online booking apps such as taxi bookings, hotels, and other types of accommodation, and parking space occupation. To put it shortly, the monitoring capabilities of traffic have merely become better and better.

The challenge we are facing is how to aggregate these vast amounts of data into powerful models that can lead to a smarter regulation of traffic that is more adaptive to different traffic situations. With good models we can get more efficient traffic reallocation in regard to construction sites, different kinds of events such as concerts or events related to sports, or weather conditions or high traffic due to paycheck week.

Overview

Our vision is to make the roads and the streets around us smarter and more aware of what is happening in terms of such events.

The success stories are already showing up. The Nevada Highway Patrol managed to reduce the number of car crashes by 17% along a stretch of Las Vegas’ busiest highway, thanks to a yearlong effort into an AI based crash prevention pilot program.

Smart traffic lights that adapt to changing traffic conditions to minimize the amount of time that cars spend idling. Through a feed of real-time data containing the number of cars present in surrounding lanes, the traffic control system may make dynamic changes in real time to avoid congestion wherever possible. In a pilot project in Pittsburg at Carnegie Mellon University they managed to reduce the amount of time motorists spent idling at lights by 40% and travel times across the city were reduced by 25%. To put it shortly, real-time management of traffic system is proven to work and increase traffic flow and efficiency. Yet these systems have been deployed on less than 1% of existing traffic signals.

Movement patterns

Monitoring of lane switching behavior and other movements

Traffic flow displays many different kinds of complex patterns that are difficult to predict. For example, bursts of traffic congestion can manifest themselves as a form of cluster formation and shock wave propagation that give rise to a stop-and-go condition for the motorists. Conditions that inhibit free flow speed of vehicles and seriously impedes traffic. The models present in current traffic flow theory are rather crude and simplified. Our neural network models may be quite a useful candidate where traditional traffic models fail.

By keeping a record of the moving patterns of vehicles, there speeds and lane switching behavior we can develop an understanding for their behavior and how different road layouts affect drivers in different situations. This is not only useful for smarter traffic control but also for identifying bottlenecks in the road layout when planning for future infrastructural improvements. Also, events such as accelerations, harsh braking and serving events can be monitored continuously and be used for accident prevention.

Geographical positioning

Real-time monitoring of accidents and other events

Today we have the technology to monitor the traffic in greater detail where we not only monitor the speed and flow of traffic but also can track movement of individual vehicles. Imagine visualizing Kerner’s three-phase traffic models in real-time on your current traffic situation. Now, that is possible.

The movement patterns of vehicles can be projected onto a map with GPS coordinates and just by looking at the distribution of vehicles, an accident situation can easily be identified quite immediately. By instant reporting of such situation, law enforcement and emergency functions can respond to incidents much faster and improve resource allocation. The system can also distinguish between cars, trucks and emergency vehicles such as police cards, ambulances and fire trucks.

Conditions in reality

Assessment of road visibility and other conditions

Even today’s weather forecasts and measurements cannot fully yield the correct picture of the actual conditions on the roads. For example, even though the temperature is freezing, the road may still be free from ice if the salt trucks have been out. Conversely, even though the temperature exceeds 2 degrees above the freezing point, there may still be ice on the road. Moreover, the weather forecasts may not be able to accurately assess the visibility conditions due to fog, solar glares or give adequate information of the amount of snow present on the traffic lanes. With flowity you have the proper technology to make such on-site assessments.

Other detection capabilities

Detection and monitoring of animal wildlife

In Sweden about 40 - 60 000 accidents involving wild animals are reported every year and the number is increasing. The reasons behind this increase is probably due to more roads being built and more people travelling on those roads. It could be argued that a more detailed awareness of the whereabouts of these animals would lead to less accidents. An alarm system that warns about mooses upon detection on the road could probably save some lives. With our technology a future were traffic accidents involving wild animals turn into a rarity can turn into a technical feasibility.

Detection of obstacles and caved in roads

With our technology, sudden obstacles can be easily detected. The event of rockfalls, landslides or caved in roads will be reported instantly where traffic can be redirected immediately in an efficient manner. Accidental spills, trailer overturns or fallen trees from hurricanes can be dealt with before posing any serious blockage to busy traffic.

Traffic anomalies and other abnormalities

When continuously monitoring traffic, you can with neural network models establish what counts as normal traffic and what counts as abnormal and deviating traffic. Along a road, a car is expected to move in a certain manner. If it deviates from the normal, perhaps this is due to an accident. The type of accident can be determined from the movement patterns of the vehicle and be reported on-the-fly for further review.