Posts

This last month has been a busy one for I-REACT. We completed the first half of the project, and presented our technology for the first time ever. But we haven’t stop there. This month we also organised two workshops: one in Boston, USA, and the other one in Incheon, South Korea.

In them, we were able to gather selected groups of professionals to discuss how modern technologies can be integrated in the fight against disasters, and the different solutions we are developing. Just so you do not miss anything, here is our brief summary of the two workshops:

I-TENDER

The I-TENDER workshop focused on how public safety services can benefit from the use of technology to respond against disasters. Claudio Rossi presented a Keynote on I-REACT project, while the other 9 presentations presented papers on how to use data analysis to filter relevant information on a disaster situation; how positioning techniques can improve the response and safety of rescue parties and emergency responders; and how technology helps in the deployment of public-safety and emergency networks.

The workshop took part in Incheon, South Korea, on December 12 and was hosted within the ACM CoNext Conference. A list of all the papers presented at the workshop can be found here, and more information on the workshop can be consulted here.

DSEM

The Data Science for Emergency Management (DSEM) was centered on the role of Big Data and Data Science in the natural hazard management area.

The keynote speaker, Prof. Carlos Castillo, presented an overview about the current state of the art on the Big Crisis Data topic and provided interesting insights on a better exploitation of crowdsourcing solutions. The DSEM workshop featured papers on novel and innovative solutions for emergency management: social media and unstructured data, crowdsourcing and user feedback, forecasting models, decision support systems, and resource allocation and crowd control during emergencies.

The workshop was held on December 11, 2017 in Boston, USA, co-located with the 2017 IEEE International Conference on Big Data. A list of all the papers presented at the workshop can be found here, and more information on the workshop can be consulted here.

We close this 2017 with a lot of activity,but we are not stopping here. This 2018 we are organizing a workshop within the ISCRAM 2018, the 15th International Conference on Information Systems for Crisis Response and Management. The workshop will be in May, but the deadline for submitting your paper is on January 15, so send it now!

ESA–Pierre Carril

Satellites do much more than to offer us astonishing photographs of space. Some of them, like the European Space Agency’s Sentinel-1 satellites, save lives. This ESA’s mission monitors the Earth day and night: it has witnessed the birth of a gigantic iceberg, helped locating the seismic fault in a Sea earthquake; or monitored glaciers in the Alps to help us understand how landslides work and helping us fight against disasters. It also provides access to remote areas, where we can’t place sensors. And the microwave radar imaging technology on board of Sentinel-1 allows scientists to have a peek of the ground through the clouds, even on the darkest nights.

The Sentinel-1 images not only allow us to have eyes in disasters that occur in remote zones, it gathers a global view of the state of the disaster with a great amount of detail. In its higher definition mode, it provides a 5 meter resolution image. The mission is composed by two satellites, which guarantees a great response speed: at European latitudes it revisits the same area about every three days. This is why this never sleeping sentry is so interesting for our project and why we needed the efforts of our partners from the Remote Sensing Research group of TU Wien. Thanks to them we will be able to incorporate the Sentinel-1 radar images into our system.

All this activity generates up to 3 Terabytes every 24 hours. This is the equivalent of creating 40 high definition movies… each day! Processing all this information can’t be done in a regular laptop. It requires the use of a supercomputer. The Remote Sensing Research group uses a supercomputerthe Vienna Scientific Clusterto analyse the Sentinel-1 data. Thanks to the computational power of this supercomputer, they are able to compare the latest Sentinel-1 images to all past data, that form a picture of a “normal looking” Earth. This allows them to create change maps that depict flooding, deforestation and other such drastic events.

Credit: ESA

But surprisingly, this daily amount of new information is not the one thing that poses the heaviest toll for computers. Establishing this “normal looking” Earth picture isn’t as easy as it sounds. Just think on the view you have through a window in your home: it changes a lot from spring through winter and from one year to another. What constitutes normal and what is out of line? To account for changes in the terrain, and the different looks that the Earth offers throughout the year, our colleagues at TU Wien are using the data already obtained by the Sentinel-1 in the past 3 years and the data obtained by the ENVISAT satellite, a past mission that went on between 2002 and 2012. Together, they constitute 1000 TB of images, more than 13000 high resolution movies. That’s certainly a lot of Netflix to catch up, so no wonder that a supercomputer is needed for analysing this huge database automatically.

Only a decade ago, this kind of technology seemed like science fiction. Today, we are integrating satellites, supercomputers and many other cyber-technologies to fight against disasters.

Augmented Reality Glasses

Over the last 5 years, Western Balkans have been severely hit by extreme flooding events. Major floods in 2010, 2013, 2014 and 2015 affected hundreds of thousands of people, causing extensive damage and a high casualty toll.

We chose this region to hold the first demonstration of our disaster management tool. In this way, we wanted to show how it can facilitate the work of authorities and civil protection in the fight against floods, but also highlight how it can aid in the coordination of different countries when a disaster hits more than one nation.

Over the three days of the workshop, the participants worked together in a simulated scenario based on the May 2014 historical Sava River flood. This flood killed 79 people, affected 2.6 million people and caused 3.8 million € in damages and losses across the Sava River Basin. The workshop linked the management of these events to the different functionalities of the I-REACT system, showing how technology can play a crucial role in the fight against disasters.

In the in-field demo of I-REACT, participants could test the crowdsourcing functionalities of the mobile app. They also tested different technologies specially devised for first responders: augmented reality glasses to provide them with live information, or a wearable that allows for detailed geolocalization.

The simulation of a control room demonstrated how a great variety of data coming from different sources, and serving different purposes, could be easily visualised by authorities. These easy visualization helps authorities making decisions in the event of an emergency.

Overall, the demo was a success. It highlighted the potential of an integration tool for disaster management, while helping authorities and responders both to save time and make more informed decisions when all variables are at play. Additionally, the I-REACT team gathered important feedback from professionals. Together with the information that we obtained in Paris, this will help us fine-tune the system and facilitate the use and integration of I-REACT within different operational procedures.

The next demonstration of our system will follow to continue bringing I-REACT closer to users and to ultimately improve future response to floods, and other disasters, mitigating their impact and help saving lives.

I-REACT group photo

The workshop was organised thanks to the support of the UNESCO Regional Bureau for Science and Culture in Europe in collaboration with the Sava River Basin Commission and other technical partners such as Deltares, the Royal Haskoning DHV from the Netherlands, the CIMA Research Foundation and ISMB (Instituto Superiore Mario Boella) from Italy.

I-REACT is an ambitious project. It relies on several technologies that must be implemented together, and that must be able to work properly, in a timely manner and under pressing circumstances. But these are only some of the technical difficulties. We face another challenge: the I-REACT system (and its app) must be easily adopted by citizens, emergency responders and decision makers across Europe. That’s why being in contact with emergency responders and civil protection agencies from the first moment is crucial for the project. Having their feedback on what their needs are is really important, so we can develop tools that are both useful and easy to use.

To gather this feedback, I-REACT relies on the experience of CSI Piemonte, and organization that has been working very closely with those who handle risks for more than 25 years. Workshops and face to face gatherings are one of the best ways to obtain information from experts. So a year ago we organized a meeting with emergency responders and other stakeholders.

At this meeting, we shared coffees with emergency responders, see them discuss together, exchange experiences… and here is what we have learnt from them, that we have been taking into account in the last year of design: their five nuggets of wisdom.

Technology is important… but implementation is key

Technology is a good ally in fighting disasters, but—as it happens with every technology—adopting new tools will have an impact on the job organization. Here’s when those coffees with emergency responders pay off: they let us peek behind the curtain, to see what impact our technologies will have on their day to day work.

Location, location, location!

This old real estate agents’ mantra can be applied also to emergency responders. We asked them about data visualization and we found that maps are the most useful tools: risk maps and road maps, to design effective strategies. We are visual beings. Hundreds of years of evolution have wired our brains to identify patterns and interpret visual information quickly. So having all the information available drawn into a map seems a perfect way to assess the situation.

All roads lead to Rome…

A while ago, we relied on very few communication channels: newspapers, TV and radio. Nowadays we have so many (Facebook, Twitter, WhatsApp,), that is hard to keep track. And there is a clear generational gap in the use of these channels: while people in their 40s onward seem to rely more on traditional channels, people under 30 use social media as their main source of information. So it is important for authorities to understand and use these new channels available, as well as the old ones.

… but some are faster than others

Over 2.5 billion people use social media channels worldwide. While this number rises every year, it is not the only advantage that social media can provide in an emergency. Social media users consume and provide real-time information, which is extremely useful for emergency responders. This communication happens almost instantly, and can happen directly between citizens and authorities, which can be a real life-saver when fighting against disasters.

Internet bots and disasters: not a good mix

In the era of fake news, where bots play on hot trends in social media, we must be able to distinguish between noise and signal. We already know that citizens can provide a lot of useful information during an emergency, but how do we separate wheat from chaff? Linguistic analysis and geolocalization will serve as preliminary sorting tools, but final decision will require the expertise of human eyes.

I-REACT aims to be emergency-responders’ eyes and hands on the ground, but we thought that the first step should be to pick their brain. This provided insights that we didn’t know before. Do you have any other insight that might be helpful? Please let us know in the comments!

“Today we can expect a 50% chance of rain…” How many times have you heard these words on the TV weather forecast? Have you ever wondered why weather-people talk about percentages? Rain, winds, temperatures… all these phenomena come with their number attached: the chance they might occur. This is the result of complex mathematical formulas of the physics behind the meteorological processes, inserted in computational models that forecasters use to predict the way weather will behave.

Since all weather forecasts models are chaotic, tiny variations on the parameters on those models lead to different results of the forecasts. These results imply different scenarios in the real life: it may rain cats and dogs, it may be a gentle rain or it may not even rain at all, but how likely is each option? In the case of rain, they combine two different factors: the confidence that it will rain someplace in the forecast area, and the percentage of that area that will receive rain if it rains. This is what meteorologist call probability of precipitation.

And, if this is important for your day-to-day forecast (so you know whether to take your umbrella or your sunglasses), imagine how important it is when we talk about extreme weather-related disasters and how to prevent them. Emergency responders and decision makers need to have at their tables all the different possible scenarios and know how likely is each one of them to happen, so they can take the best possible decisions. That is why at I-REACT we are including weather-related data and models into our I-REACTOR, the system that will integrate this information altogether with satellite and UAVs images, crowdsourced information and many other data sources and technologies, to provide detailed disaster risk maps for Europe.

Forecasting extreme events (like high levels of precipitations or strong winds) is key for preventing disasters. And to do this, special forecasts, different from the weather forecasts you see on TV, must be designed. Our colleagues at the Finnish Meteorological Institute are in charge of providing the extreme weather-related data. This means that they feed different extreme weather scenarios to the system, each one of them accompanied by a number: the chance that that particular scenario may happen. By doing so, FMI is able to provide different thresholds for risks: a probability that may seem tiny for normal events can be of huge importance when associated with extreme weather events.

Instead of delivering a unique weather prediction, FMI runs several simulations with slightly different initial conditions, so we can know the different scenarios and know how likely is each one of them to happen. This is called the Ensemble method. To calculate these different scenarios, FMI uses complex numerical models that run on supercomputers. The accuracy of those models depends highly on the initial conditions: the starting points of the simulation, consisting on real data taken from satellite images, meteorological stations and other sources.

To provide the most reliable results, FMI is feeding their models the best available data at the moment: high resolution maps gathered from weather systems across Europe, with a resolution down to 7 kilometres on a European scale, and a 3 Km resolution on a national scale. a much higher resolution in comparison with the usual map in you TV weather forecast which resolutions usually goes down to 20 km.

By combining better resolution maps and more accurate probabilities for extreme events, I-REACT will be of great help in saving lives thanks to cutting-edge technological advances. Against disasters, we have a fighting chance. And now, we are better at calculating these chances.

It was not by chance that drones became the gadget of the year in 2014. By then, thousands of units were being sold worldwide with prices ranging 30€ to 30,000€ and big electronic retailers were reporting huge surges in sales. Now, a few years down the line, this trend has continued in the rise and drones are a relatively common leisure for taking pictures and shooting impressive aerial videos. However, did you know they can also be used to fight disasters?

After Typhoon Haiyan hit the Philippines in 2013, an unprecedented number of drones patrolled the skies to aid in the humanitarian response. Thanks to this, a much rapid mapping of the affected areas was possible, which was essential to setting up humanitarian base camps, detecting the most affected communities, locating victims in need of help, and assessing the state of infrastructures for transportation, among others.

The importance of drones in disasters is that they provide a quick view from above of the affected zone. For decades, this information was only obtainable using planes for aerial photos, or satellites, which have a number of limitations including cost, data sharing restrictions, cloud cover, and the time needed to acquire images. In contrast, UAVs can provide a bird’s eye perspective quickly, if people are at site, and at a far higher resolution and at much lower cost.

An additional key aspect is that drones can also offer a view that is not perpendicular to the ground. For instance, when assessing damage after the disaster, whereas satellite vision allows seeing if a building has a roof but not if it has four walls, oblique vision from drones can answer this question and many others that require more three-dimensional information.

But overall, the biggest advantage is that, unlike satellites, citizens can own UAVs and this means that disaster-affected communities can participate in response to a crisis. Now, after major disasters such as the Philippines typhoon, but also the Haiti earthquake, these countries have taken the lead in involving citizens and there is a growing number of grassroots initiatives to teach local people to operate their own drones in emergencies.

At the I-REACT project we believe in the potential of citizen participation in disaster response at European level and we are developing an app that will allow connecting emergency professionals and drone users. Our partner AnsuR is leading this task to allow decision makers at control centres to request and conduct flights over affected areas, both from amateur and professional drones, in case of floods, fires and extreme weather events. This ways, they will be able to assess and ensure the relevance of the images while drones are still flying.

In order facilitate the engagement of volunteers from local areas, AnsuR is building a database for drone volunteers who potentially can be involved in the event of an emergency in their area of operation, and creating a efficient system for communicating the images they capture. This way, any European citizen from local communities can become an integral part of the fight against disasters, improving rescue operations and protecting their communities.

If you own a drone and would like to contribute, drop us a line and join us in quest for a more resilient future!

A trend in technology is the one that does its job without you even noticing. This is the basis of wearables, whose ever expanding list of application ranges from capturing body parameters to sleep quality. But wearables are now much more than just leisure gadgets: their ability to report positioning and sense the environment can bring about a revolution to many professional fields. This is especially true for disaster response.

Although the forefather of wearable technology is a miniature Chinese abacus ring from the XVII th century used for mathematical calculations, the first popular reference are wristwatches. Since the early XX century they dominated the market and evolved into different uses until they were progressively replaced by cellphones and other more recent wearables. Consumer market is flooded with devices equipped with different sensors for activity tracking, heart rate or oxygen level monitoring. There are literally thousands of very low-cost fitness, health or wellness-oriented wearable devices. But also, among the mainstream markets there are various medical and assisted living devices for continuous health monitoring. Wearable technology thus helps the growing number of patients overloading hospitals and medical centers, while increasing their life quality.

Today the most distinctive feature of wearable technologies is the ability of exchanging data without human intervention thanks to electronic sensors and new firmware/software. This passive gathering of data is particularly crucial in the event of an emergency where the capacity of professionals for acting is very limited by time. In this scenario, capturing information about the status of the environment (e.g. temperature, air quality etc.) and of the rescue teams (location, activity, vital signs etc.) can be critical.

Wearables are commonly used in disaster response with examples like the wristband of Morphix technologies for the detection of hazardous chemicals, among others. However, the technology is not exploited in its full potential. As part of I-REACT, the Serbian company Bitgear is in charge of the development of a wearable for first responders that will be the first device applied to disasters with both positioning and sensing capabilities.

3D representation of the I-REACT wearable

The advanced navigation technology provides a much more accurate position than regular GPS, that can have high deviations of accuracy and large positioning errors in urban environments. For this, Bitgear is using a multi-constellation receiver which combines raw satellite navigation data not only from American GPS but from the European Galileo/EGNOS and Russian GLONASS. The integration of different sources with the processing of raw data through algorithms and coupling with the inertial sensors (INS) provides a much reliable positioning than any portable device used in disasters nowadays. Bitgear is also working towards expanding the initial device concept to state-of-the art real time location system (RTLS) that will combine Ultra-wide band radio (UWB) to provide indoor positioning of the rescue teams at critical situations.

The functionality of environmental sensing will be used for the detection of risky scenarios for first responders. For instance, if the oxygen level drops only four percentage points from the standard level (21%), this can impair coordination and judgement of the rescue teams. Thus, anticipating this environmental changes is essential. Also, by assessing the drops in oxygen levels we can obtain another relevant information as they might indirectly indicate the increase of toxic gasses. Thus, with the I-REACT wearable, rescue teams will know when they need to wear masks when necessary to prevent poisoning.

The I-REACT wearables will be connected via low-energy bluetooth to the mobile app developed in the context of the project. This way, the sensing and positioning will be sent to the big data structure and readily provided to decision makers at control centres.

For the implementation of this technology there are a number of challenges. On the technical side, the design of electronic devices for harsh environments, such as those found in emergencies, is always complex and requires good materials and insulation. Another challenge is posed by the proximity to the human body since it absorbs electromagnetic energy, which degrades the signal of the device. So the materials, the position of the antenna, the topology of the electronics have to be tweaked. Finally, one of the most important issues is to build the smallest possible device to avoid overloading of responders that are already forced to carry many gadgets. To this end the miniaturisation process will be very centred in the efficient placement of oxygen sensors, as these are usually very bulky, and minimization of obstruction of radio signals.

To date, different functional wearable prototypes have been produced and they are in the process of performance evaluation and environmental testing. Also, different options for boxing are under development. All in all, the device should be ready by the end of this year.

The application of wearables to the I-REACT project holds the promise of a safer and more effective coordination of rescue teams, and demonstrates that overall technologies are an essential ally to fight disasters.

When Hurricane Katrina hit the American coast in 2005 Facebook was a newcomer to a still-to-be-developed world wide web, there was no Twitter to have news updates and less than 70% of citizens owned a mobile phone. Today, with more portable devices than citizens and an ever-constant interaction through social networks, the way we obtain and share information during crisis has drastically improved. This is proving very helpful in recent crisis like the 2013 super typhoon Haiyan in the Philippines where Twitter was the single greatest information source for response and recovery efforts.

Social media is becoming essential for authorities to access vital information provided by citizens that would not be available otherwise, which improves the prevention and response to critical events. However, social network information is largely unstructured arising from the fact that everyone can be an information source. From eyewitnesses to emergency responders or NGOs, that can provide information from the ground, to mass media that amplifies the message, or even outsiders showing sympathy and emotional support. In this context, there are many factors that affect how the information flows, such as the use of hashtags which is very diverse and can sometimes hamper the identification of relevant data. Thus, it is necessary to analyse social media to place the pieces of the puzzle together.

The extraction and analysis of social media information is an important part within the I-REACT project. This information obtained from citizens will complement data coming from earth observations, UAVs, or emergency responders, among others, to provide real time data on floods, wildfires, earthquakes and other natural disasters. For this, Natural Language Processing (NLP) technologies developed by the I-REACT partner CELI, are being used to analyse big data streams from social media.

To do this, great amounts of information are initially collected from social networks by using searches on generic keywords such as “earthquake” or “flood”. Although this information will be unstructured, all or most of the emergency-related material will be gathered this way. Since this data can be compared to that of past events and to “regular” behaviours on social networks, a vital information will be generated: detecting if something unexpected is going on and spotting the occurrence of an emergency in real time.

This information will then be validated through linguistic analysis and machine learning techniques. Here, it is possible to select the emergency-related contents and identify useful information such as the type and location of event, the casualties, or the damage to infrastructures and services. In addition, we can also have information about the sentiment of the message, which is important to create panic maps and to prioritise actions on the ground. And once the event is concluded, the system keeps collecting data so that it can be continuously tested in spotting new emergencies from social media. This way, this tool will progressively learn and refine its ability to identify disasters.

Overall, social media analysis provides fast and relevant information during emergencies, highlighting the fact that these communication channels are not only changing the way we live and interact with each other, but also making every citizen an essential part in the fight against disasters.