Creating clarity in today’s complex Early Warning Systems
The research with which the UT has contributed to the interoperability of emergency assistance systems is called 'Interoperable Situation-Aware IoT-Based Early Warning System' and was carried out by research assistant João Luiz Rebelo Moreira. It is part of a large European research project, Inter-IoT, which aims to design, implement and test a framework for the interoperability of various Internet of Things platforms.
‘Early Warning Systems are information systems that use the Internet of Things to monitor changes in the physical world and to send off alerts whenever changes occur that may involve risk. These and other Internet of Things appoications are becoming more and more complex’, explains project leader Marten van Sinderen. ‘The data these systems use are becoming increasingly diverse: they are collected data from very different domains, sent to varying target groups through different systems, each of which has its own communication protocols and data formats. The challenge is to integrate all those data flows as quickly and as carefully as possible. That’s where interoperability comes in. The aim is to ensure efficient exchanging and combining of data, and a rapid, clear presentation of the integrated data without loss of information.’
The tension in these efforts is evident, notes Van Sinderen. ‘We’re always looking for the optimal balance between semantic quality and representation efficiency. Semantic quality is about giving meaning to the data: what value do they add? Representation efficiency is all about the cost of communicating, processing and storing the data.’
Van Sinderen explains that information processing for emergency service involves five steps: collection of data; detection of risks (for example, on the basis of deviation from normal conditions); decision making; informing; and responding. ‘There can be a lot of noise and loss of accuracy between the moment the data are collected and the moment of response, or intervention. You can imagine how that might affect a disaster situation, in which emergency services have to cope with photo and video images, satellite images, weather data, personal data of those affected, traffic information, and much more. The trick is to get all those data – or as much of them as possible – to the emergency workers as clearly, quickly and unambiguously as possible. The better that works, the faster and more effective the intervention will be.’
Part of the UT project was a case in the port of Valencia, Spain, where data on truck logistics and truck driver health were successfully integrated in one model. Van Sinderen: ‘Placing these disparate data within a consistent framework facilitates decision-making and interventions. It also allows users to see the causes and consequences of incidents at a single glance. For example, take a collision in which a driver goes into cardiac arrest: access to the data shows you what happened first, the collision or the cardiac arrest. This is just a simple example of how data from different sources and areas can offer more insight into a situation. Obviously, this makes prevention easier and more effective as well: as soon as you recognize patterns in what is happening, you can take measures to improve.’
Another important result of the UT’s research is that it yielded an extension of the so-called SAREF standard (SAREF stands for Smart Appliances REFerence ontology). This standard is used as a frame of reference for processing data in smart appliances, such as devices connected through the Internet of Things. ‘The extension, SAREF4Health, is aimed at monitoring the health of individuals. Like the successful case in Valencia, the standard extension is a small, but important step forward,’ says Van Sinderen. ‘We’re taking evolution, not revolution. Yet the implications are far-reaching. In our world we are increasingly reliant on complex information systems – in business, industry, traffic, the police and justice. Interoperability means that all the available data can be used in the best way possible for better decision-making and more effective prevention.’