/ AI

How BCaster can help with disaster response

Before joining BCaster at the end of 2017 I had worked in disaster management departments of the International Federation of Red Cross and Red Crescent Societies (IFRC) and Finnish Red Cross for nine years. During my time in the IFRC’s Asia Pacific Disaster Management Unit in particular, we were always trying to figure out new ways to gather reliable information from the disaster-affected areas as fast as humanly possible.

Humanitarian organizations can deploy even field hospitals capable of complex surgical operations within 24 hours after a disaster has struck. In addition to speed, it is critically important to send the right people and the right type and number of relief items to make sure the affected people’s needs are met. What good does it do to provide shelter kits to people if the earthquake didn’t damage any buildings? There’s of course always local and remote assessment teams dispatched to the scene to assess the needs but, in disaster response, time is of the essence and getting any useful information as early as possible from the epicentre is always helpful.

There’s already a plethora of excellent initiatives on harvesting and collating user-generated content (UGC) for supporting disaster response operations, but there are still a number of unsolved issues the BCaster technology can address:

  1. Lack of location metadata and the inability to search by location. When you look at different media platforms, only a small subset of them provide location-based search and even then a lot of the media content is missing the location information altogether. Capturing videos and photos with the BCaster app or any BCaster SDK-powered app ensures you always have the correct time and location metadata associated with your media. It’s then a breeze to search for content from a specific place over a specific period of time on our platform.

  2. The struggle to authenticate media has existed ever since photography was invented. We all know of sad stories where even major news organizations have unsuspectingly used media with fake content or altered metadata. We employ multiple technical means to ensure what you see really did happen at the time and place shown on our platform.

  3. Defining the right search words is hit and miss at best: on most platforms users are responsible for adding the description, hashtags or keywords, which leads to multiple issues, namely a) the lack of descriptive keywords in the first place, b) synonymous keywords (think typhoon vs. hurricane vs. cyclone vs. tornado), c) using local language for keywords (“how do you spell earthquake in Nepali again?”) and lastly d) the inevitable delay caused by users having to think about the best description and keywords for their media before publishing. On BCaster the keywords are assigned automatically to all uploaded media with the help of our cutting-edge AI. Try uploading an image of a damaged building or a flood scene to the BCaster platform and you will always get “rubble” and “flood” keywords assigned to them automatically every time - in a consistent manner.

  4. Upload delay - On most platforms uploading content always requires some kind of user interaction causing mild to severe delays before the user-generated content is searchable and accessible. On BCaster the content is uploaded automatically after capture. (You can also choose to upload via Wi-Fi only if you’re on a limited data plan.) If you capture in public mode, the content is available for search almost in real-time after the recording stops. Natural disasters can of course take their toll on mobile networks but networks are becoming increasingly resilient due to technological advances.

Combine all of the above and you have a very powerful end-to-end media pipeline in your hands. For the humanitarian sector this could allow decision-makers to act faster with greater confidence in those early hours after a sudden-onset disaster.

For a news company it could mean tapping into a constant stream of authentic crowdsourced UGC with content-, time- and location-specific alerts and flexible models for monetization and rewarding content creators.

For a professional sports team it could mean driving up fan engagement with two-way communication and providing sponsors detailed visibility analytics with the help of our game-changing AI solutions.

We can integrate into existing apps and media pipelines or offer just the AI analysis for your media if you have the other parts covered. I would love to discuss in more detail if you find value in these features, whether you’re working in the humanitarian or private sector.

Best wishes,
Eero Sario
Software Development Lead
BCaster Ltd.