Many of our readers will be familiar with how Google can use our device sensors to determine where we are without resorting to the power hungry GPS / GLONASS chipset. Google have used crowdsourced data from cell towers and WiFi routers for a number of years as these report their location. For many applications and services, this approximate location is good enough and saves battery power. However, it is not just Google that can crowdsource data and not just for location services. UK company OpenSignal have been using an application to crowdsource data from Android (and more recently, iOS now that Apple have invented multitasking) users for some time now. Launched in 2010, OpenSignal have used their data to help customers know the signal and network speeds. OpenSignal have already started branching out to use other uses of more crowdsourced data and released a report this week that explained the technology of using mobile devices for crowdsourced data was “one of the most exciting technological innovations of recent times…. Mobile sensor networks present the opportunity to gather data about the world at a never-before-seen scale and low-cost.” OpenSignal, Waze, Google, Sensorly and Rootmetrics are great examples of crowdsourced data managers.
OpenSignal’s core application and service tracks users’ cell ‘phone coverage and makes this data available to anyone. It’s great if you want to compare carriers in given locations as you won’t have to believe their marketing departments. Because it relies on users installing the application (and providing their data) OpenSignal offers customers a number of additional features such as a best coverage compass (points you in the direction to head if you need better speeds), local public WiFi hotspots, a network ranking tool for given locations and a data transfer speed test. OpenSignal have other ideas about using handset sensors for additional functions: a ‘phone’s barometer, designed to improve location accuracy, may be repurposed to report back the atmospheric pressure. Light sensors, designed to tell the device how bright to make the screen, can be used to estimate sunlight intensity. The raw data must be processed in order to be useful and things are not yet perfect. There are many ideas, for example the color of light could also be measured to determine if the smartphone is indoors or outdoors. One side effect of this processing is that the algorithms make this information very context aware. This context awareness will help make our applications smarter and more useful, although the uses go much beyond this simple idea. Much bigger ideas and projects can benefit hugely from crowdsourced data.
Crowdsourced data may be used for a large number of things on both a local, regional, national and global scale. When we put the data into the context of the Internet of Things and how many services in our lives are currently independent from one another, but how if there was the right information (and a compatible means of communication), we could make life easier and kinder to resources. Let me put things into perspective: crowdsourced data could be used to estimate the weather in one region. It could be compared with forecasts, processed, and used to help smart thermostats and energy providers determine if it would be beneficial to heat or cool houses in anticipation of a change in temperature, using cheaper energy at times of low demand.
I’d like to go back to OpenSignal. They’re a business and have currently made money by selling network coverage reports to carriers, but with the developments to their application they may soon be in a position to be able to sell on data to other companies. For example, OpenSignal’s application has been installed on Android-based devices all over the world. This means that it has access to users of forked versions of Android, such as Amazon Fire devices; this in turn means that captured data is broader than that obtained by Google alone (as Google uses Google Services for this information). This in turn means that OpenSignal are able to produce fragmentation reports based on their user database. However, their data can go at least one step further and may be used to provide information based on the components used in devices: what processor chipset, version of software, hardware networks and similar? This information could be very valuable to a developer considering writing an application that requires a certain hardware component for a given region: it will give them an idea of how many devices would support the application.
OpenSignal has also started to launch standalone applications with specialized roles, such as CrisisSignal; this is a realtime version of OpenSignal designed for supporting disaster aid workers. It was tested after Hurricane Sandy in 2012 to pinpoint where the strongest networks connections were to help find the best location to set up aid centres. It’s being used as part of the Ebola crisis in West Africa. But this is just the beginning! OpenSignal appear to have quite the lead in innovative thinking with regard to context aware data from our devices. If you already use the application, let us know what you think to the services that it provides. If you don’t, are you going to consider giving it a try? Leave us a comment below.