Artificial Intelligence on the M-30 applied to road innovation

Due to the introduction of artificial intelligence on the M30, systems capable of developing intelligent behaviour have been created which facilitate management and maintenance tasks on the road. This is one of the reasons why the M30 has become an international benchmark in the field of innovation.

AI has brought important advantages, especially in error reduction, difficult exploration, daily application and the start-up of digital assistants.

Artificial Intelligence Systems on the M-30

Thanks to the application of artificial intelligence on the M30, systems have been developed that help measure traffic in real time and detect anomalies in lighting or water pumping wells. In this way it will be possible to act and solve the incidents of the road with greater immediacy.

Real Time Traffic Evaluation System (SETTRE)

A total of 367 traffic measurement points have been installed to be able to receive data in terms of speed, intensity and real-time traffic occupancy of the M30. Analyzing this data produces:

• Identification of incidents and consequent action: with the data collected and depending on the density of these measurement points, the possible incidents will be identified in order to act accordingly as quickly as possible.
• Calculation of travel time on the M30: In addition, the SETTRE system also provides information on the travel data of each of the journeys along the M30. This information helps us calculate the travel time on the M30 routes from the Emesa website.

Automatic Lights Off Detection Algorithm (LOADA)

In this case, the algorithm  detects and counts the illuminated screens that remain off and the Maintenance department is notified to solve it as soon as possible.

It is activated once a day and analyzes each of the 667 images that are collected along the entirety of the road and branches of the M30. For each image, the algorithm first detects the situation of the line of lights, analyzing which lamps are turned off by looking at the neighbouring lights. And secondly, looking for the discontinuities, with a screen turned off for each one.

What’s the result? A new image that, based on the original, shows the detections:

• In red when a discontinuity arises.
• In green if discontinuities have not been detected.

 Water pumping wells

We remind you that the M30 tunnels run on both sides of the Manzanares River. Therefore, deep drains have been placed through pipes that feed 100 pumps distributed via 32 pumping wells.

We carry out an exhaustive surveillance on each pump, with which we collect the start or stop state and the time at which the change of state occurs.

The data we obtain helps us to find out the sequence of starts in each well and establish patterns in the starts of every pump. Thus allowing us to draw the following conclusions:

• The time a pump is started up and the duration.
• If it has started its turn according to the sequence in the well.
• If there has been a significant deviation from the others in the well. This data may be indicative of a loss of performance and, therefore, susceptible to being reviewed.
• Adapting a model of this behaviour of every pump and well is done instantaneously through information from the last two months. Additionally, every minute an analysis of the most recent behaviour with regards to modeling is carried out.

Road innovation has facilitated obtaining relevant data in technological solutions to improve services on roads as relevant as on the M30. Emesa uses artificial intelligence technologies on the M30: SETTRE, LOADA, water pumping wells, and a long list of different processes to help decision making.