Two Italian scientists tested how accurate smartphones are in detecting earthquakes. This test is not just about technology, but also about new business models at the nexus of mobile, the Internet of Things and big data. Executives, in environmental agencies, emergency services, academia and other public sector entities in charge of monitoring earthquakes, have to figure out how this new paradigm will affect their organization.
My dear country Italy is blessed with a lot of natural wonders: mountains, beaches, lakes, and rocky coastlines. But it packs up quite a few risks of natural disasters too, such as avalanches, landslides, floods, volcanoes and, last but not least, earthquakes. So no wonder we have experts in the field of earthquake research. Two of them recently published an interesting paper in the Bulletin of the Seismological Society of America. Without getting into the scientific details, which I do not even dare to understand, they compared traditional seismometers accuracy with micro‐electro‐mechanical system (MEMS) accelerometers embedded in smartphones - in particular they tested the LIS331DLH MEMS accelerometer installed in the iPhone - in terms of ability to provide earthquake observations. The conclusion of the paper was that "The main drawback of the LIS331DLH MEMS accelerometer is its low sensitivity, due to the high level of instrumental self noise, and so it can be used effectively only to record moderate to strong earthquakes (ML>5) near the epicentral area". So, in layman terms, good, but not yet comparable with seismometers buried in purpose-built wells, or buildings, bridges and other man-made structures. Still a fascinating development, right at the nexus of mobile, the Internet of Things and Big Data. Think of it, millions of smartphones recording observations that can be sent real-time to a command center to alert emergency services, even in areas where seismometers are not installed. Data that can then be stored for further forensic analysis.
To better grasp what the technology and business implications for government environmental agencies and emergency services could be, I tried to look at a practical example - admittedly an overly simplified one. Lombardia, Veneto and Emilia-Romagna, three of the largest and wealthiest regions in Italy, set up an earthquake monitoring network with traditional stations. From what I can understand they have installed approximately 30 seismometers, mostly in concrete wells, that transmit observations to a central system, although some of them are inactive. Considering that each seismometer can cost between €10K and €20K, depending on the technology (those buried deep in oceans cost much more, but not sure smartphones would be of any use there!), we are talking €300K of investment in equipment. And we have not even began to consider the construction and maintenance of concrete wells, the power supply and the dedicated fixed line network to transmit data to the central system.
Now let's make some assumptions about smartphones. I am using a conservative estimate here. Eurostat says that the percentage of individuals that used Internet applications (apps) on a handheld device exploiting the information about where they are (e.g. GPS) was 7%, in 2012. One can safely assume that those individuals are a sensible proxy for citizens that own a smartphone with an embedded MEMS and that they would be willing to use an earthquake monitoring application, if someone made it available for free. If one applies that percentage to the population of Lombardia, Veneto and Emilia-Romagna, which amounts to approximately 19 million citizens, it would get to 1.3 million citizens that could be sending observations about earthquakes. Almost certainly more observations that the one collected through the 30 or so seismometers, at no cost for devices, concrete wells and power supply. But it's not just the cost, it is a change of paradigm in the business model. Instead of being a governmental program controlled by experts in concrete wells and seismometers, earthquakes monitoring becomes a big data problem, where data scientists and CIOs play a much bigger role. The challenges would not be about supplying power, installing seismometers, avoiding that wells get flooded or damaged. They would be about figuring out how to consolidate data that come from different apps, developed upon iOS, or Android, or Windows mobile operating systems. It would be about the continuity of operations of wireless networks. It would be about developing an app that works even when the phone is off (earthquakes have the habit of taking place at the strangest times, such as at night, when people are asleep and their phones are off). But I'm sure that would not be a problem, my phone's alarm clock does it every morning… unfortunately. It would be about storing many more data points than for traditional earthquake monitoring networks and putting them to good use, for real-time alerts and predictive simulation for contingency planning.
The two eminent Italian scientists that wrote the aforementioned paper can probably do a better prediction than I can, to judge how farfetched the scenario I depicted is. My prediction is that smartphones will just be a complement for seismometers for a few years. But it is important for environmental agencies, emergency services, universities and other public sector entities involved in the business of predicting and responding to earthquakes to start to think of this change in paradigm.