When disasters strike, edge computing must kick in

Gathering data for situational awareness should still take place during natural disasters, even if there’s no internet, say scientists. Edge computing must perform the task.

When disasters strike, edge computing networks must kick in
REUTERS/Nick Oxford

Edge computing and fog networks must be programmed to kick in when the internet fails during disasters, a scientific research team says. That way, emergency managers can draw on impacted civilians’ location data, social networking images and tweets and use them to gain situational awareness of scenes.

Routers, mobile phones and other devices should continue to collect social sensor data during these events, but instead of first attempting to send it through to traditional cloud-based depositories operated by the social network — which are unavailable due to the outage — the geo-distributed devices should divert the data to local edge computing, fog nodes and other hardened resources. Emergency officials can then access it.

+ Also on Network World: What is edge computing and how it’s changing the network +

It’s “a new way of gathering and sharing information during natural disasters that does not rely on the internet,” Georgia Institute of Technology says in a media release.

“Using computing power built into mobile phones, routers and other hardware to create a network, emergency managers and first responders will be able to share and act on information gathered from people impacted by hurricanes, tornados, floods and other disasters,” it says.

Data-vacuuming smartphones are now ubiquitous. The problem, though, is no internet means no informative data science to manipulate or on-scene-describing tweets to get studied in the respective emergency operations centers.

The researchers' idea, however, is to offload this highly useful intelligence locally so that it can be turned into real-time human density maps of inundated flooded areas, for example.

Edge computing mediums

The school cites surveillance cameras, embedded pavement sensors and smartphones with increasing amounts of processing power as being suitable edge-processing mediums.

We've seen how mobile network operators (MNO) are taking advantage of edge computing themselves. It’s used to reduce latency. Those phone companies are increasingly using local computing boxes (often inside their many buildings, left over from the days of copper-requiring phone switches, and on their towers) to store and process data rather than centralizing it.

“This ability will give a huge advantage to first responders,” Georgia tech says of its idea. The team of researchers published a paper (pdf) where they describe their “fog-enabled social sensing services” API.

In the paper, the researchers describe how docker-friendly fog nodes connect or relay the distributed social sensors — the smartphone-carrying civilians, in other words — to hardened routers that can perform edge data processing and be pinged locally, or a swarm of low-power Raspberry Pi mini-computer-loaded drones could produce high-resolution images for a flying ad hoc network “and provide a completely new level of mobile computing.” All takes place in a geographic area that doesn't have a functioning infrastructure.

One key to making it all work, the researchers say, is to get the social sensors to perform the edge computing — the aggregation of information and filtering lowers the data overhead requirements. That’s something a powerful smartphone should be able to do.

And the cloud is still involved — it runs updates, setup and so on for the components when internet is available.

“We believe fog computing can become a potent enabler of decentralized, local social sensing services,” the researchers say.

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.
Related:
Must read: 10 new UI features coming to Windows 10