Blog AnalysisConsumptionCivil protection

The ecall system, the panacea for saving lives? Or a big lie for a good industrial lobby

Systemeecall logo(BRUXELLES2) A marvelous system which, in the event of a car accident, calls the ambulances on its own and saves lives. This, in a nutshell, is how the system was presented e call that the European Union wants to install on board all new vehicles and voted these days by the European Parliament. The reality is slightly different...

Five Gross Mistakes

I called upon my memory, my experience, my archives, and a few experts. We discover, in fact, a device that aims more to satisfy the objective of a connected car than to facilitate emergency rescue. A project more supported by the automotive industry, especially telecoms for that matter, than by the emergency services. The argument of saving lives is widely put forward: 10% less road deaths according to the European Parliament! But it turns out, in fact, to be very largely inflated, not to say "candy". A simple examination of "common sense" makes it possible to realize this. It reveals at least 5 gross or methodological errors (read below). The disadvantages — overloading the emergency call centers — are minimized. The real stakes of road safety — the prevention of road accidents, the quality of alerts and first aid, the effectiveness of emergency relief — are completely overlooked.

A project born from industry

This project was born in the early 2000s on the one hand in Adac-type assistance companies in Germany (1). He is then conceptualized as the GSM guardian angel (read here). And for almost 15 years, there has been frenzied lobbying, especially by companies from northern Europe through the Ertico project, which brings together automotive and connection companies (Volvo, Nokia, etc.). On the other hand, the Peugeot PSA company has implemented a system in its top-of-the-range vehicles, since the mid-2000s, which allows it to be interconnected with assistance companies (such as IMA - intermutuelle assistance)

Supported by DG Telecom

Within the Commission, this project was not carried out by DG Health or the Civil Protection unit (today grouped with humanitarian aid), nor by DG Transport, nor even by DG Industry but by DG Information Society (DG Connect today), within the framework of thedigital diary, and the "intelligent cars" program! For more details, this site also refers to an "ecall box" site, which refers to a site apparently of general interest,, funded by the European research program, which turns out to be in fact the site supported and supplied by... Ertico.

Imperatives far removed from emergency services

Suffice to say that this project has only a very distant relationship with the effectiveness of emergency relief, but more with the connected car market, and secondarily with car assistance companies. The objective is also that this device is found within an assistance package - managed by these companies which could also be the first call center of the vehicle and ensure the necessary sorting. In fact, it is also based on a device and a technology that is found, partly outdated, by all the mobile devices currently deployed (GSM, smartphone application, etc.).

How many people saved? An estimate... very overestimated!

The new device could reduce by about 10% per year” the number of road deaths - 25 people in the EU in 700 - or 2014 lives saved according to the European Parliament press release. Marvellous ! How not to say applaud with both hands... The reality is that this figure is false. It's taken from a European Commission impact study that doesn't really say that. A study itself subject to caution because it is based on presuppositions, which seem anything but scientific and inaccurate.

The original sentence: With eCall, emergency service response time could be reduced by 50% in rural areas and 40% in urban areas, leading to a 2-10% reduction in mortality and serious injuries between 2% and 15%, depending on the country! (NB: population density, road infrastructure, emergency response systems)

1st mistake: take the top of the range (and not the average)

The figure presented publicly therefore only takes the top of the range (10%), if all vehicles are equipped and for certain countries: either by 2038. It does not take into account the bottom of the range, or even an average option: the figure of 4 to 6% is often used in studies.

 2nd mistake: taking into account all road safety

This percentage is then calculated on all road fatalities, all road accidents combined, including those not involving a vehicle or not allowing the system to be triggered. However, just under half of fatal accidents (45%) affect the most vulnerable: pedestrians (22% of deaths in Europe according to European statistics), bicycles (8% of deaths), even motorcyclists (15% of mortality). It is not obvious that this system is triggered (sensitivity identical to that of the airbag), for an impact which can be light for one vehicle and fatal for the other (pedestrian, motorcycle, bicycle). Sometimes a simple deviation from the vehicle can cause a neighboring motorcyclist to skid, without even a shock.

3rd mistake: extrapolating from national studies

The emergency time reduction figure, taken up everywhere afterwards, comes, in fact, from a German study, "Storm". A study initiated in the Stutgart region by the German automotive manufacturer Mercedes Benz, as part of a project called PROMETHEUS supported by the European research program (Eureka) and the Land of Baden Würtenberg, carried out from 1992 to 1995, which aimed to improve... road traffic management. This figure was then extrapolated and generalized at European level, with no other verification than empirical (with a wet finger!).

The German study "STORM" (Stuttgart Transport Operation by Regional Management), showed almost 50% rescue time improvement in rural area, with a net gain of around 10 minutes. Rescue time in urban area would be improved by 40

4th error: halving the intervention of the emergency services

This sentence is a monumental error. There, we are no longer on the approximation but on a gross falsehood. We extrapolate the gain on the alert time (where the ecall system essentially intervenes) to the overall time before medical treatment. However, the alert time (on which the ecall system intervenes) is only a small part (approximately 10%) of the overall emergency response time (less than 18 minutes in France for 90% of interventions) . And it seems very difficult to count on such a drastic reduction in the overall intervention time.

Alert scheme - source: BSPP
Paris firefighters alert plan - in 6 steps - arrival at the hospital is missing - source: BSPP

Explanation. An intervention time is generally broken down into 3-4 elements:

  1. The alert. In general, it is made by witnesses (the generalization of mobile phones has changed the game, making it possible to greatly shorten this alert time) at a dispatching center (112 in Europe, 15 or 18 in France). NB: the study made by the company PSA mentioned 50% of calls within an average time of less than 1 minute, 12% between 1 and 2 minutes and only 2,5% beyond 2 minutes (NB the remaining quarter did not answer). (3)
  2. The time taken by the emergency services to react — between the reception of the call at the dispatching center and the departure of first aid —: this depends both on the responsiveness of the person in charge of receiving the call, on the clarity of the alert but also on the telephone congestion of the alert center, and the availability of the most suitable means (steps 3 to 5 on the attached diagram). NB: In France, a study by SDIS - fire and rescue services - mentions an average reaction time of 1'56" (with a help 2'40" maximum for 90% of calls).
  3. Time to get there for first aid (medicated or not). NB: In France, this same study mentions an average delay of 10 minutes for traffic accidents and victim rescue (with a delay of 13'43" for 90% of interventions). In Paris, a hyper-urbanized area, the rigorous time for the Paris Fire Brigade (military) who provide most of the interventions is 7' on average, and 12' for the Samu.
  4. Medical care, either on site where possible, or in a hospital setting (4).

If we add up these average times, we arrive at a support by the emergency services within 13 minutes (18 minutes for most of the interventions). We do not really see how this overall intervention time could be reduced by almost half thanks to "ecall". It's magic! It should be added to this that in an urban environment, or on a highway/national road type road, where a good half of the accidents occur, the "warning time" factor is very relative. It is more the duration of the trip that will count - traffic congestion and traffic jams - especially in urban areas delaying the arrival of an ambulance. Similarly, what matters is having a vehicle capable of getting there quickly.

In fact, the study mentioned by Ertico and the European Commission starts from an average delay of 21' outside built-up areas, including an average delay of 2'2 for detection, 7'8 for communication, 1' for notification and 1' for availability and departure, and an average response time of 9'2. System StatEcallIntervention@Ertico

What lends itself to very strong questioning is the time - extraordinary of almost 8 minutes for communication (which seems exaggerated). As for the average reduction in the intervention time of 1 minute, it seems totally extraordinary at a time when GPS is widespread, including in emergency services (generalization which did not in fact exist in the early 1990s!) . Suffice to say that this figure is truly bogus!

5th mistake: disregarding other tried and tested calling methods

According to the result of a French study, carried out by the industrialist PSA, this system is especially effective in a “rural area, at night and alone in question” configuration. It's there that " assistance by the emergency services can be the most delayed and the most difficult: difficulties of alert and intervention, localization ". The study, which concerns 300 cases, 40 of which were particularly detailed, notes that in 8 cases out of 10, there were other witnesses on the spot (3).

With the generalization of mobile phones - which has greatly changed the situation and shortened the call time - even smartphone applications - which can further shorten it -, and allow you to also call an emergency or assistance device directly, not to mention the generalization of GPS, the "Ecall" device risks finding itself quickly outdated.

How many victims really?

If we put all the data in order, we arrive at much less miraculous results. A little calculation I did. If we take an average option of 5% efficiency (very generous assumption), on the basis of 60% of fatal accidents, vehicle accidents + certain motorcycle/bicycle accidents, instead of 100% (also fairly generous assumption) , and a relative efficiency of the system (20% where there is no other means of appeal), we only arrive at a potential reduction in mortality of only a few dozen people throughout Europe (120 to 150 people in an optimistic hypothesis, or reduced to the scale of France, about fifteen dead). On the scale of a road safety campaign, over one year, where we can observe a gain of 1% in mortality (in bad years!), this is very very low. And again, we do not take into account the medical reality of roadside assistance: the chance of saving a victim is decided in the first minutes. Suffice to say that we are more faced with a gadget than a real emergency device.

The perverse effects of the eCall system

This device has several perverse effects little highlighted by the European authorities (and the manufacturers who promote it).

One is often put forward: the number of false calls that will be generated. Accidents with material damage, causing a major shock will cause a series of calls, which will encumber the emergency services with calls that are perhaps necessary but not really urgent. The first thing to do when you have a material or even minor accident is to get out of the vehicle - and not to disconnect the system!

The other is less mentioned but just as notable: to overcome the previous drawback, this system will be connected to a central alert which, itself, will manage the dispatch of help. It's pretty smart. This limits the risk of false calls to the emergency services. But, on the merits, this adds a level to the scale of the help (listening time, then passage of the alert to the competent service) and amounts to destroying to almost zero the gain in speed hoped for. Another disadvantage: we do not say how much this additional device will cost the car driver.

Finally, third aspect, very little highlighted: this device requires a telephone connection, type GSM (mobile relay). In "white" areas - which have no connection - the ecall system will not work. Because there will simply be no possible connection.

What is this system really for?

In fact, this system is especially effective in a few specific types of accident: roads with little traffic or uninhabited areas, mainly at night, single vehicle involved, with single person(s) on board and/or semi-conscious (s). Suffice to say that we are not in the generality, and even in the tiny minority (except in a few countries, notably the Nordics, which have already set up a system of this type).

In the vast majority of other accidents (whether material damage or involving a victim), this system has a direct interest (as for the remote alarm at home): to allow a person who is the victim of an accident - who is not necessarily deadly - to come into direct contact with his assistance. This allows him to have human contact, allowing him to "de-stress" (if necessary to "unload" his stress) and to identify his needs fairly quickly. This is certainly a very useful function, but it has more to do with assistance than with emergency rescue, and in any case very little to do with road safety.

This "ecall" device will also make it possible to trace vehicles from the point of view of driver behavior (or even serve as a "snitch" for insurance) or even for the purpose of combating crime or delinquency. Officially, every precaution is taken to keep the "box" placed in the vehicles its "emergency" ability. You can have serious doubts. Because no guarantee is really provided. We will see in 10 years what are the "real" uses of this "box".

The real reduction in road fatalities

If we want to tackle road fatalities or make rescue more effective, this requires well-known actions, no doubt less spectacular than Ecall but far more effective.

Act on the causes

Stat SecuriteRoad Evol15ansThe real reduction in mortality on the roads is based on a fight against the main causes: alcohol and speed, in the first place, the state of the roads also or the aptitude of the drivers.

It is by acting on these elements that road mortality has been able to decrease significantly in recent years. In 15 years (since the ecall project was launched) road mortality from vehicle accidents has been reduced by more than 3 in France (-322%). It dropped significantly despite the increase in traffic.

A "good" year of road safety sees mortality decrease by 5 to 10%! As much as the estimated, and very eventual, gain under the most optimistic (and unrealistic) assumption of the ecall system at a 20-year development scale. (By the way, this underlines the very relative effectiveness of this system in terms of road safety as a whole).

2014 road safety report in France (ONISR January 2015)
2014 road safety report in France (ONISR January 2015)

Mortality could decrease further. Without a doubt. But at a rate probably lower than that experienced in recent years. The 0 accident does not exist. And the solutions to reduce mortality are now more or less known: reduction in speed, driving under the influence of alcohol and toxicology. This presupposes, in large part, both a strengthening of controls and prevention campaigns. It is based on the protection of "fragile" people (pedestrians, bicycles, motorbikes). We are far from "ecall" technology.

Download :

The real question of emergency relief: training and money!

The "real" question of emergency relief and the reduction of mortality is the effectiveness of the emergency system, its good distribution throughout the territory, the availability of modern means. However, today with the budgetary tightening, the means of help are calculated as accurately as possible. This limitation concerns, first of all, the call reception staff, which will be limited to the medium emergency level — 1 or 2 people in the small centers (or at night), 3 to 4 in the others, or even a little more in very large centres. It also concerns the means of intervention.

A good emergency regulator will therefore "save" expensive medical or rapid means (helicopters, motorcyclist support), in the most serious cases and where there is a clear hope of being able to save a life. E-call system or not. Ditto at the hospital level: if we take the Paris region, at night, there is only one duty scanner! What is called the "great guard". This for reasons of economy. If we really want to give back more, we must densify the rescue network, and make more use of medical helicopters in less urbanized areas. This is the reverse of the sought-after path of savings.

The other way to obtain a rapid intervention, at a lower cost, is to have a maximum of people trained in first aid, who can very easily give the alert and first aid. Because nothing replaces human intelligence on automatic devices. This presupposes giving a new impetus to rescue training both in schools (colleges, high schools, universities) - at the time when driving is learned - and in the granting of driving licenses. It's less spectacular than an ecall system but surely more effective...

(Nicolas Gros-Verheyde)

(1) Proof of this interest can be found on the website of the European Commission, which is content to illustrate its point by reproducing a diagram of the ADAC

(2) Virtanen, A., Schirokoff, J., Luoma and R. Kulmala (2006) Impacts of an automatic emergency call system on accident consequences, Ministry of Transport and Communications Finland Finnish R&D Program on Real-Time Transport Information AINO - Download Finland study 2006


(3) Download:  experimental study PSA 2002


(4) Even in the presence of the best of the medical teams on the spot, in a short time, only a treatment in hospital environment will allow to undertake the operations which will allow the survival of the patient

Nicolas Gros Verheyde

Chief editor of the B2 site. Graduated in European law from the University of Paris I Pantheon Sorbonne and listener to the 65th session of the IHEDN (Institut des Hautes Etudes de la Défense Nationale. Journalist since 1989, founded B2 - Bruxelles2 in 2008. EU/NATO correspondent in Brussels for Sud-Ouest (previously West-France and France-Soir).