Data and connected objects serving the safety of populations

by bold-lichterman

New technologies must take their place in the safety of populations, at 3 key points:

BEFORE: It is possible to better predict crimes and attacks before they happen.

WHILE : React better at the time of an attack, at the scene of the attack

AFTER: Better find the culprits, speed up investigations and arrests.

Algorithms based on the “Big-data + Connected Objects” pair will be the most effective, but require the installation of connected sensors in our cities.

Technologies to prevent crime BEFORE it happens

In the movie Minority Report (2002), crimes were “prevented” and prevented even before they were committed, with the help of technology. In the USA, Palantir and Forensic Logic, two Californian companies, make predictions of terrorist crimes from data available on all past terrorist actions. In the United States, the Future Crimes Institute brings together technical security experts to help police forces understand the risks and imagine such solutions. The Big Data platform Kaggle recently allowed the best computer scientists on the planet to analyze a set of San Francisco City Security Data in order to develop this type of predictive algorithm.

Unpremeditated crimes, such as crimes of passion, will fall through the cracks of algorithms, but according to Goodman, a former police official and adviser to Interpol, “Big-data analysis has for several years been able to predict many crimes. premeditated such as car thefts and street crimes, thefts or homicides. ” So when a “number comes out” in the TV series Person of Interest, or a computer algorithm, the police will have to decide how to use that information. Should we ask a judge to order reinforced surveillance of a person who is designated “only” by an algorithm? In the absence of evidence, the police will have to figure out how to fairly use this statistical information without infringing the rights of honest citizens.

The data does not always show the whole situation and the algorithms decide according to their setting of the moment… Street lighting is one of the good weapons of prevention. The more a street is lit, the more secure it is. Do drones need to illuminate sensitive areas at certain times, according to data from a big data crime risk analysis?

The data includes suspects’ DNA, facial recognition information, or suspicious cash withdrawals from cash machines equipped with cameras, which would-be terrorists often use to receive or transfer money from, where to, abroad. On the one hand, the Bitcoin is an effective counter to remaining anonymous in money exchanges, just as “Tor” allows you to buy weapons anonymously.

On the other hand, Coulibaly had taken out a loan from Cofidis to finance his attack.

Technologies to minimize crime WHILE it happens

The streets of major cities around the world will probably see the appearance of even more CCTV cameras, including thermal or night vision cameras, as they help speed up police investigations. Depending on the level of insecurity, these devices will be increasingly accepted, or even demanded, by the populations. The cameras of American society FLIR are currently among the most efficient.

Respect for individual freedoms

However, technologies must make it possible to secure public places without alienating individual freedoms.

Security gates

It is – for example – conceivable to place security gates at the entrance of all public places only on condition that people are informed (as they are in compliance with the posting rules set by the CNIL ) but the passage must still be done “without realizing it or slowing down” so as not to remind everyone that the world is less secure than before. These controls must be “invisible and fluid”.

Historical security actors such as Morpho Detection have a role to play. Why is there not yet a revolving door including a metal or explosives detector? This type of “natural-looking” detection solution would erase the anxiety-provoking aspect for populations.

The analysis of video surveillance images can be observed by algorithms, by employees or by… the population. Why not anonymously broadcast images from CCTV cameras on the Internet by changing cameras regularly to allow everyone to monitor their own city, country or planet in a collaborative manner on the principle “You watch my streets, I watch them? yours ”.

This hypothesis had already been launched for images of airport baggage scans, but had been blocked by public opinion in its time. Baggage checks have not yet been able to take advantage of the rise of the collaborative economy, to the detriment of security. Are the French ready to accept the principle of “crowd-watching” video surveillance images? There is no point in shouting “Big Brother” since we give back to citizens control over images of life in the streets.

Sound detection

Sound intensity sensors can allow the automatic recognition of particular noises such as submachine gun fire, which would trigger an automatic emergency alert. No need for a human to take the risk of raising the alarm in front of a machine gun!

Explosives detection

Can we replace explosives detection dogs with connected objects – an invisible and non-intrusive device – for analyzing the air quality in public places?

Fingerprint detection

We are all equipped with Smartphones, some models of which can take fingerprints. Fingerprints have already been hacked and falsified. It is therefore not the ideal solution.


When a crime is committed, the assault must be able to be preceded by a flight of drones. These drones can be equipped at will, depending on the purpose of the mission:

– Cameras to take photos of attackers for accelerated facial recognition,

– Weapons on board the drone, to stop criminals, such as tear gas drones to immobilize the aggressor,

– Dazzling drones

The drones-ambulances tested in the Netherlands to provide first aid would also save time in rescuing a greater number of victims, by relying on the “local workforce” already present on the scene.

A French company had wanted to offer first aid “motorcycle ambulances”, in order to bypass traffic, but this has – unfortunately for the victims – never been installed.


Far from the constraining “electronic bracelet”, geolocation technologies have largely developed with the rise of smartphones. Should we consider permanently geolocating all the persons listed “S”? This raises concerns for individual freedoms, and must therefore be very well regulated legally. We could thus detect suspicious and unusual movements, or the meeting of certain suspicious individuals among themselves.

The geolocation of suspects’ smartphones can already be continuously monitored, in order to answer the following questions: What are the changes in movement compared to the usual? What are the abnormal movements? What are the exceptional encounters between people “followed” by their phones?

These are tracks, “traces”, which are left by all connected objects used by people under surveillance.

Geolocation of vehicles

Should we require the possibility of geolocation of all cars in order to be able to locate, or even stop automatically, all stolen vehicles?

Automated video license plate analysis technologies would also make it possible to monitor the movement of suspicious vehicles – similar to the principle used in London for the payment of tolls for access to the city.

Information sharing between police forces is also essential. In Europe, this is undoubtedly only possible with the establishment of a common European political government and a European Ministry of the Interior. It will come in the next few years, hopefully soon.

Without sharing data, how can we track suspects from one city to another, from one state to another, from one country to another. Terrorism is global. Some of the terrorists in the Paris attack of November 13 were in Belgium the day before.

The Department of Crime from the University of Pennsylvania (USA) which has developed its own algorithm to predict the next victims of homicides from a wide variety of data.

Social networks also provide a lot of data to analyze. Algorithms now know how to detect suspicious behaviors, keywords or exchanges.

On November 12, the account of “_jacky_boy_ “ posted on Instagram (by amateurism or by mythomania?) “When you see this photo on BFM, it will be too late.”

Technologies to manage the situation AFTER the crime

The arrival of help

Some countries offer the possibility for fire trucks and police forces to remotely control all red lights in their path, in order to turn them systematically green.

In a few years, the arrival of self-driving cars could greatly facilitate the arrival of help on site, by sending a request to modify the route of the cars located on the road.

Crimes are often caused by repeat offenders.

Should we use algorithms to assist sentencing judges in their judgment of the dangerousness of prisoners to be released, or not, before the end of their sentence?

In the United States, this is already possible in more than 80% of cases, with the systematic analysis of 24 data: the prisoner’s file, the age at which the crimes were committed, and this would have made it possible to reduce the rate of 15% recurrence.

Follow-up after release from prison

Should all people leaving prison be systematically condemned to recording their travel data, for the sole purpose of subsequent consultation in the event of an investigation? I can already hear the cries of the guarantors of freedoms, and they are right. But it is technically possible.

There is a fine line between individual liberties and security, and technology allows the cursor to be moved to one side or the other. The decision on the cursor position will come – I hope – from the people themselves.