Finally, a report conducted by a study group of competent people full of common sense, from the French senate, has recognized the uselessness of the SAIP mobile application which has been developed under the previous French government as a way to warn the population about terrorist attacks in their neighborhood.

What is the French SAIP mobile app project?

The SAIP mobile app project (Système d’alerte et d’information des populations) was ordered by the previous government as a response to the multiple terrorist attacks that struck France in 2015 and 2016. The government contracted a private company “Deveryware” with a budget of 300,000 Euro to develop a mobile application which would send alert notifications to smartphone users in the vicinity of a terrorist threat.

Why there was a need?

Clearly there is a need. Multiple terrorist attacks in Paris and Nice created a situation of paranoia of the people, who have been traumatized by the events, when the communication of critical information would have helped the population to react appropriately, hiding or avoiding certain areas.

The only public alert system in place in France is called “Réseau National d’Alerte” or RNA. It consists of a network of around 4,500 electro-mechanic sirens that has been deployed after the second world war in all French cities, which will play a specific modulated sound 3x during 1 minutes and 41 seconds.

The system is also tested the first Wednesday of each month with a 1 different modulation of 1 minute (source test sound).

The RNA sirens are interconnected with the old PSTN lines of Orange, via an infrastructure that is barely maintained since 2013.

This system is outdated and ineffective, as a study showed that the great majority of the people (78%) does not know what to do if those sirens would are activated, while it will cost more than 80 millions Euros to keep maintain and renovate in the next 3 years.

Why the SAIP was a stupid idea?

Coming up with the idea of using a mobile app to broadcast public warning messages shows the lack of understanding of both telecommunication networks and the mobile app world.

1. It will reach only the people who (i) have a smartphone AND (ii) have downloaded the app AND (iii) have not uninstalled it. Given that people change their phone in average every 2 years, the challenge is even more complex to enure (iv) they install the app again on their new phone every time they change, considering that the app will be useless for them 99% of the time.

2. It will require the deployment and maintenance of an infrastructure for message broadcasting with very high availability. There are plenty of solutions which can be used, mostly based on MQTT (source). These solutions require a service to be running on the phone at all time (which consumes more battery and is an additional cause for uninstallation by the user). Furthermore, the notification traffic is treated the same way by the network as any of the rest of the traffic, whether it is a Facebook feed or pictures of cats.

3. International standardization bodies such as 3GPP and ETSI have developed standards to implement high availability, resilient Public Warning Systems (PWS) in mobile networks with all the mechanics to push a broadcast information to the population of a designated area. It relies on Cell Broadcast service which forces a messages on the handset of all users in a specific coverage area. It works on most of the phones today, on 2G, 3G and 4G networks, and is not that complex to implement.

What now?

Scrapping the budget of the legacy sirens system to interconnect the government to mobile networks to and implement Cell Broadcast is a move in the right direction. In the US, the Amber Alert is using heavily Cell Broadcasting to inform the population about wanted persons and vehicle details for cases of kidnapping.