Modern cities nowadays are characterized by the omnipresence of IT systems used for supervision, citizen services and management that support the city everyday operations, producing a large amount of data. Recently this volume of data has sky-rocketed, mostly because of the sheer volume of real time environment generated data (cameras, sensors) as well as the user generated content (mostly from social media, blogs, RSS feeds,which cities both produce and consume).

 

The same nature of the video, audio and social data, being continuous feeds (compared to the traditional documents which until recently made up the majority of the city’s generated content) makes their assimilation and exploitation at the different levels of the city hierarchy difficult. Thus currently feeds are only consumed when produced (i.e. monitoring the cameras) and have very little value to other decision systems up in the hierarchy.

 

However existing multimedia search solutions do not provide effective search over arbitrarily large and diverse sources of multimedia data derived from the physical world.

 

The solution

"SMART" project is constructed according to a modular system architecture for search and retrieval of multimedia data from multiple sensors. Intelligent techniques for real-time processing, search and data retrieval from the environment have been developed as well. Reasoning techniques are applied on the data retrieved, to extract meaningful results. Linked open data are also used when available.

 

In more detail, the solution offered by SMART is composed of:

• The SMART Social Network Manager queries the virtual world of social networks

• The SMART Search Engine which uses statistical analysis to search and rank results

 

Furthermore the various stakeholders in the value chain of the city operation (municipal administration, citizens, visitors, private contractors), who have different interests in the information produced, are offered different data visualization options, according to their needs, in the formof ready to use dashboards.

 

Re use of existing infrastructure

The platform will re-use and protect the investment of the city, both in terms of already installed cameras, existing weather stations or any other type of sensor as well as in terms of existing network infrastructure. If no such infrastructure is in place, readily available, low cost off-the-shelf equipment may be used.

 

Practical use

Following installation of cameras, sensors and data validation activities, the platform may be used to identify and distribute interesting content in realtime, effectively supporting services such as

• Locating where the action is right now (information from cameras and audio)

• Identifying the busiest open air hotspots (using data from traffic sensors, parking activity, headcount, noise)

• Hearing the sounds of the city

• Monitoring what people say about the city (using social media analysis)

• Identifying potential security threats (identify abnormal activity in regularly calm areas of the city)

• Providing city live information (e.g. multisite concerts, live events, etc) to create syndicated feeds (especially interesting for tourism and culture)