Mediaboss Marketing

Were halfway through 2020 and its safe to say that the year will be forever associated with the COVID-19 pandemic. For most of the last six months travel restrictions, lockdowns, social distancing and mask wearing have been put in place in an attempt to curtail the spread. Now, as the world attempts to reopen, governments and businesses must deal with the dual problems of restarting after the shutdowns and, in the absence of a vaccine, trying to contain further outbreaks. Technology is playing a key role in the research trying to understand the virus and eventually developing that vaccine. Can it also play a part in containing the spread until it is deployed? Can it help life return to normal?

Addfor, an Italian company with almost two decades of experience in Artificial Intelligence Solutions development for engineering, has created an AI-enabled system called citySAFE that can help to monitor so-called hot spots as well as mitigation efforts in an attempt to limit the spread. citySAFE interfaces with already installed camera systems or mobile cams, to provide aggregate information for either outdoor or indoor public and private spaces. Lenovo is collaborating with Addfor by defining and providing the right system hardware configuration to support the immense raw processing power required by a high camera-count video streaming.

The citySAFE application relies on a simplified interface that displays critical situations in real-time, leveraging available data about infection and hospitalization rates, then color-coding areas accordingly. For example, if a hot spot springs up, citySAFE can, in real time, assess the extent to which mitigation efforts in that area are being followed. All indices, such as counting populations, population density, or percentage of mask usage are calculated in real-time and grouped in space and time to be explored both geographically and temporally with an advanced graphical interface. Health officials can then issue warnings and deploy personnel to remind people and businesses about masks social distancing, and other safe practices. This same scenario could play out for a private company with a large footprint campus, manufacturing or distribution complex.

citySAFE was recently tested on a large scale in Turin, Italy with the collaboration of the public administration and other local authorities. It is the only integrated system available at the moment that allows the timely and continuous monitoring of an entire city, either from city surveillance or aerial cameras, for the issues related to the control of the spread of the COVID-19 virus. By using existing city camera infrastructure, citizens are subjected only to their existing level of CCTV observation.

This type of continuous streaming of live video data requires tremendous processing power, both at the edge, where the video is captured, and in the data center where the results are compiled, run through the algorithms for inference, distributed and retrained. In this context, Lenovo is providing the unique rugged ThinkSystem SE350 edge server. By using the computational power of the high-performance NVIDIA T4 GPU, the ThinkSystem SE350 delivers video streaming wirelessly from cameras, and real-time inference at the edge. On the back end, the Lenovo GPU dense ThinkSystem SR670, designed to support up to eight high-performance NVIDIA V100 GPUs, is built for running large AI workloads such as citySAFE and scales linearly as requirements grow. Both the ThinkSystem SE350 and SR670 are built on Intel Xeon processors for optimal performance and security.

We know that the pandemic will not go on forever: A vaccine will be found and deployed. Life will return to normal. When it does, citySAFE and its associated infrastructure can be repurposed for developing new or enhancing existing services the city administration or private companies provide citizens or employees such in areas such as safety, parking management, waiting times, and advanced digital signage systems for traffic or missing persons alerts.

In 2100s, future historians will study the COVID-19 pandemic, the same way that weve studied the Spanish Flu outbreak of 1918. They will see the mistakes from 1918, some corrected, some repeated, in 2020. And they will see a global crisis fought on a local basis. Mitigating the spread of the virus in the interim until a vaccine is developed, while restarting the economies of the world, is the tightrope we find ourselves on now. How we execute those precarious steps may end up being the yardstick by which those future historians measure us. In the face of these challenges, how did we get life to return to normal?

Go here to see the original:
Using Artificial Intelligence in the Fight Against COVID-19 - HPCwire