PEOPLE | Mar 2, 2017

The Km4City experience: interview with Paolo Nesi

How to improve quality of life through the creation of innovative services based on data in the Km4City experience

Created in 2013 as an ontology for smart cities, Km4City aims to improve the quality of life by enabling the creation of innovative services for mobility, security, tourism and reduction of consumption and costs, and by enhancing the ability of people and the city to react to unexpected adverse events.

These are ambitious goals that can be achieved by creating services using artificial intelligence algorithms based on data present in our cities. Not only open data, but also those of operators in the fields of transport, trade, tourism, cultural heritage, education, weather and environment; those that can come from sensors in the city (IOT, the Internet of Things), from social media and from citizens themselves through App and systems of participation.

In addition to static data that do not change, or rarely change, there are dynamic data (real time) that produce continuous streams of information and arouse greater interest for end users because they make it possible to obtain information, forecasts and deductions in real time.

As a knowledge base at the service of various projects, Km4City currently covers the whole of Tuscany with its data in terms of road information, points of interest (culture, tourism, accommodation, catering, education and business for a total of about 300,000 POIs classified into 500 categories), public transport (of 16 operators), petrol stations, information on hospital triage, traffic flows from about 800 sensors, over 200 car parks and social media via Twitter Vigilance; hundreds of thousands of new complex data a day.

We spoke about the project with Paolo Nesichair of the DISIT Lab at the University of Florence and project coordinator.

Who are the beneficiaries of the Km4City project?

The main beneficiaries are the city’s users (citizens, students, commuters and tourists) and, of course, operators and public administrations.

Km4City is an example of how intelligent use of data can support the City in planning and developing new services. Can it be replicated in other areas?

Km4City is completely open source, it can be easily replicated in other regions and areas. It is also a completely open system to which new data and processes can be easily added. The ontology itself is open like everything else. The server side systems have been developed in open languages such as Java, PHP, Javascript, ETL, Python, etc. The Km4City platform is modular and scalable, in the sense that a small administration can also install and use only the parts that are of interest to it. The entire data acquisition system is able to ingest/aggregate static and real-time data in multiple standards with solutions that are fully scalable in the cloud or in the data centers public administrations.

How is Km4City fed?

Km4City is fed with open data, road graphs and real-time data from open data, mobility managers, social networks such as Twitter, from any type of sensor in the city and also from apps and thus from the end users themselves.

Why is Km4City a benchmark in the international arena?

It is not up to us to say that Km4City is an international benchmark. The ontology and thus the Km4City model have been judged the most complete by various bodies. In addition, Km4City solutions are being used by various European and national projects that also contribute to their expansion as models and tools. These projects include: Sii-Mobility (Ministry of Education national smart city, mobility and transport, with several trials in almost all of Tuscany), RESOLUTE H2020 of the European Commission (for the study and implementation of resilience solutions for transport systems with trials in Florence and Athens), REPLICATE H2020 of the European Commission (for the introduction of sustainable mobility solutions, IOT and integrated energy, in the context of the European smart city plan), the Ministry of Education’s GHOST project … and others.

Can Km4City become an important asset in support of urban planning?

Of course, given that with its algorithms it is able to produce origin-destination matrices on vehicle flows and also on people (with the contribution of RESOLUTE), and create predictive models on flows, parking and people. It collects data and knowledge on how users use the city and on how the city lives and evolves. On the basis of knowledge, tools are being activated for decision support, risk assessment and the evaluation of resilience.

What critical issues have been encountered?

The main issues that have been addressed and resolved are related to management of the complexity of heterogeneous data acquisition (different sources, protocols, standards, formats, etc.), to the volumes of these data arriving continuously from the City and from Apps, and to their quality and discontinuity; in this way configuring the smart city issue within the space of Big Data.The lack of interoperability and the limited quality of data are handled in Km4City through data mining tools for aggregating data and correcting problems within acceptable parameters. On the basis of Km4City knowledge, a variety of data analytic algorithms are then executed that are based on artificial intelligence and statistics for the production, also in real time, of forecasts, suggestions, stimuli for citizens, and suggestions for policymakers.

The next steps?

We will be engaged in the expansion of trials with Sii-Mobility to many areas of Tuscany on aspects of mobility and transport, and to Florence and Athens on resilience aspects of transport systems. In detailing data, based on the requests we receive, we are expanding the knowledge base in order to cover more regions up to reaching the whole of Italy.