Marseille, Capital of surveillance
Between 1,500 and 2,000 surveillance cameras will be operating in public areas by 2021.
1/8 of the city’s police force are assigned to the Urban Surveillance Centre (CSU) which operates 24/7 – that’s 47 police officers.
The cameras are essentially concentrated in the most central, commercial and touristic areas. The neighbourhoods reputed to be more crime-sensitive are under-resourced. A political branding tactic?
- Installation: €20-25,000
- Rental costs: €290 to €870 per month for the entire duration of its life.
- Average cost of operation: €4,600 per year (price including rental and maintenance).
17 million euros
This is the overall investment cost of the first phase of video surveillance deployment (2012-2016).
46 million euros
This is what the 500 new cameras have/will cost (2018-2021).
In use, the system of ‘video protection’ costs the city around 7 million euros each year.
In 2016, 90% of the Urban Surveillance Centre’s activities was issuing sanctions for offences caught on video
In 2015, the system of video surveillance provided useful image evidence to around 1% of investigations led by the national police.
But surely surveillance cameras have a deterrent effect and lower crime in this way?
No, there have been no studies that show a correlation between video surveillance and falls in crime.
Big Data Observatory of Public Tranquillity.
In 2016, the city of Marseille announced a budget of 10 to 12 million euros to develop ‘Smart Video-protection’
Controlled by a private operator!
Ineo – subsidiary of Engie.
The company won the bid for a sum of 1.8 million euros.
The specifications evoke the possibility of ‘detecting the preparation of rogue events or delinquent acts’ (e.g. large convivial gatherings, illegal marches, racketeering around schools), and ‘anticipating risks/threats likely to derogate public order and citizens’ safety (e.g. evaluating the risk of dangerous gatherings by analysing tweets)’.
All that, ‘based on the identification of involved parties (who speaks? who acts? who interacts with whom?), and searching back through conversation feeds (who is organising? who is the first to act?)
This is a digital mechanism of automated surveillance of the city and the population, produced from a large amount of data. Its launch was initially scheduled for the end of 2020.
The mechanism allows the analysis and corroboration of data coming from the fire brigade, hospitals, the public transport network, but also from social media, telecommunication operators, and from citizens themselves.
In response to the weak results of video surveillance, a faction of the political class and security entrepreneurs proposes Automated Video Protection.
Facial and biometric recognition, automatic image analysis, predictive policing, and all that.
In 2018, the market for the deployment of an automated video surveillance mechanism is assigned to the company SNEF, an integrator of video surveillance solutions based in Marseille. The city has not communicated on the actual scope of the project.
Automated video surveillance.
What is it?
The automatic processing of data (…) so as to detect anomalies/incidents/out-of-the-ordinary events.
The “detection of non-identifiable anomalies by an operator” and the “management of public spaces, analysis of pedestrians/vehicles as well as behaviour”.
The detection of “abandoned objects”, of “graffiti tags” and of “theft/disappearance/destruction of urban furnishings”.
The search of images “aided by filters”, one of which being “individual (description, avatar, photo)”.
The project specifies that the city council reserves the possibility of adding certain functions, such as “sound detection” (explosions, gunshots), the “reconstitution of events” (like the journey/route of an individual), the detection of “abnormal behaviour” (fighting, stealing, aggression).