The collection and analysis of large volumes of data(big data)for the resolution of problems such as road traffic is obviously necessary, but not enough. The researchers work in the creation of machines capable of thinking by themselves and, without the intervention of man, provide solutions to new problems that arise.
This was explained yesterday by the associate professor of the University of the Basque Country and researcher,Javier del Ser,during his speech on the eighth day of the Atlantic Logistics and Transportation Exhibition (SALT 2018), globally entitled Smart Mobility: New trends.
During his lecture –Intelligent mobility and big data: new advances from Data Science and optimization– Professor Del Ser pointed out the importance of a moment like the current one in which, unlike what happened when he began to investigate in this field,yes there is data."There is a substrate", summarized the present capacity to observe reality and extract common patterns that allow to establish solutions.
Now the "gardener" is needed, he continued, to water and grow the plants. For example, the nearly 5,000 vehicles that are concentrated every Friday between 2:00 pm and 3:00 pm in the hub ofTorre Las Palmasof the Gran Canaria capital, demand some type of solution.
Established this, the models and algorithms pursued by researchers would allow the machines to respond by themselves to any eventuality: a greater volume of trucks for the existence of works, the closure of a lane … And what would be the answer? A change in the cadence of the traffic lights could be one of them.
In a word, that the machines have that"good cubero eye"learned from the experience that presides a good part of human decision-making before facts that have never occurred before.
In the opinion of this telecommunications engineer doubly doctorate – "I'm not a weirdo," he said -, the Science of Data "applied to mobility has to be adaptive and as autonomous as possible." So far, thebig dataIt has fulfilled a descriptive function – it indicates where, when and what size are the traffic jams – and predictive – there are going to be traffic jams – and now look for theprescriptive,offering alternative routes, for example, before the evidence that the volume of vehicles will generate problems at one point.
That is why the machines must be able "to learn and update themselves before the arrival of new data". For this, the figures that nourish computers must have volume, speed, variety and accuracy. To those four uves, Del Ser added two more:value and variability.
"We look fordata aimed at establishing an advantage for citizens ",he explained. And, in addition, work is being done to create "non-static models, if the work is changed, the mobile phone has to adapt to the new situation efficiently", he continued to exemplify this characteristic that the machines in which the present.
It may sound like cinematographic fiction, but the concretion seems closer if the summary capacity of the professor of theUniversity of the Basque Country:"However complex the system, it can always be reduced to a mathematical problem."
To achieve such an ambitious goal, a change of the third in theinterinstitutional cooperation.Javier del Ser gave the time passed when the advances were jealously guarded under seven keys, hiding them from the rest of the researchers, potential plagiarists. Even so, he acknowledged that there is still "competition and isolation of groups, even within the same company".
These attitudes constitute "a problem in the face of offering solutions to society". According to its description, the profile of those working in this field must gather characteristics such ascuriosity, creativity, ambition and tenacity,understood as the ability to "fight with a problem day and night until the solution is found, and someone tenacious can not be put in a sealed compartment". To break these borders he created the Joint Research Lab, "a cage of crickets – I am the biggest cricket -", in which information circulates freely.