We are living in a world where AI, data, and ML have occupied the world completely. And the world we are living in, and our knowledge have been put on a test with the Coronavirus appearance. It tests us how ready we are for it, and whether it is a surprise or just a lack of readiness and usage of technology to fight it. This article is based on our latest webinar on using AI to fight pandemics, and outlines several points:
- Pandemics in the past and their spread
- Importance of technology and prompt usage of data
- Pandemics level and properly used data
- Use of AI to fight pandemics
1. Pandemics in the past and the way they spread
If we look back in history, we can see that pandemics and flu exist even from the beginning of our time. Data shows that even in the past have also been many death cases, which were maybe even worse than COVID-19, but the only difference with the situation now is that in the past, social media didn’t exist to spread the word so fast. COVID-19 for that matter appeared at the end in 2019, in China, and it seems that even though the information is fast and widely spread nowadays, we still didn’t prepare ourselves good for it. We didn’t quite leverage the power of the information that we had before COVID-19 closed us in our homes, stopped our parties, traveling, and life in general. But, if we compare the way COVID-19 spreads, we can see that measles, chickenpox, rubella spread much faster than COVID-19. Regardless of this, the mortality rate of COVID-19 is a bit higher, and it is really dangerous for humans, companies, countries, and the economy as a whole. In the picture below, you can see a timeline of the pandemic outbreaks throughout history, as well as a representation of COVID-19 spread and other pandemics.
2. Importance of technology and prompt usage of data
Though technology has advanced a big time, it doesn’t seem that we leveraged it to fight COVID-19. And we could have done that, we could have leveraged Big Data, IoT, BlockChain, and Artificial Intelligence. You might wonder how? Well, it is simple. We are living in a world that is interconnected, we have a fast Internet connection and various possibilities to collect data from various sources. If we properly collected that data, used the data from the IoT sensors, the security that Blockchain offers, and the Artificial Intelligence to process, and analyze that data, we could have made a predictive model that would fight COVID-19. But we reacted slowly. The reason for this is that technology is developing exponentially, and the organizations, governments, people are still not adjusted to the possibilities that it offers.
Closely related to this is the huge amount of data that is produced on a daily level. There has been research in 2019 about the amount of data that is being produced within 60 seconds, and the numbers are astonishingly huge. And that was in 2019, now we are 2020, so imagine that the number is probably doubled. COVID-19 has increased this number even more – social media platforms are full of information about new cases, fighting the disease, preventive measures, etc. We read and hear about this every 10 minutes, if not less, so the number is probably tripled. But, regardless of all that data, it doesn’t seem that we employ it much in practice. How can we leverage it?
When it comes to using data, one of the most important things is the reaction latency. It means nothing just seeing the data, but acting upon it – then that is something which we should pay attention to. For example, in the past when the Black plague occurred, people started using data to discover the types of infections that happened, and though we have those records, we didn’t employ them in practice. Additionally, when the first cases in China appeared, we were ignorant of all that data, and if we have used everything that we have heard, then we could have prevented the situation from going worse.
3. Pandemics Level and properly used data
If we analyze the pandemics, we can see that they have effects on three levels: company level, national level, and global level.
a) Company level
COVID-19 closed all companies – we are all working from home and adjusting to the new working environment. This poses challenges both to companies and employees. Companies have to make sure that their employees can perform the job, are well-equipped, and in a good mood. Employees have to deal with the challenges of burnouts, the missing social component, distraction, and lack of working context. To ease the job, many of the companies have created applications for connecting with their employees, sharing the latest companies’ policy updates, following their employees’ health, sending tips & tricks for a healthy working environment, etc. And all these applications are based on the data provided by employers and employees.
b) National level
There are a lot of tools that can be used to analyze the data and perform predictions on a global level. One example is taking the data and presenting it in dashboards of where the infected people are located, the number of infected people, deaths, age, quarantined people, etc. All this information will give us better insights as to how fast the infection is spreading and what we should do. Below you can see an image of one analysis that we have done based on the data that we had. If you want to see the entire analysis we have made, based on the data for North Macedonia, download the full document here.
c) Global level
Having a pandemic on a global level is a really serious matter. We see that from the situation that we are in. A few years back there was an attempt from Google to predict the flu in the USA. However, that was an unsuccessful attempt as they didn’t gather proper data. So, looking back on that attempt, the situation that we are in, the technology, and the knowledge that we have, what we need to do is a crisis analysis.
As mentioned at the beginning we have various data sources. So, first, we need to gather all that data, from our mobile phones, internet, open data, sensors, etc. Look at all the challenges that come with it, such as overestimation, noisy data, false data. Take into consideration the future directions regarding the reliability of data, and analytics, disaster tolerance, etc. and create a Central WHO Data Warehouse. This global warehouse would contain data of all countries regarding population, the density of population, movement, health data, etc. Then, epidemiologists, ML & DS experts, researchers, states, universities, and private companies can jointly work on creating Machine Learning and Artificial Intelligence models and simulations to prevent pandemics or influence the future outbreaks.
If we were ready for the COVID-19, we could have taken the data about the settings, and countries, compared them and then we could have made predictions about how the virus would spread within a country. Having this kind of data would have made it easier for everyone to employ preventive measures and stop the virus spreading around the globe.
4. Use of AI to fight pandemics
Once we have gone over the history of pandemics, Data, and pandemics level we are getting to the point of how we can use AI to fight pandemics. Many people think that AI is nothing more but just a plugin that is inserted into the computer and can solve everything. But, AI is much more. It is, in fact, the sum of all the senses that humans have, and with AI we try to mimic our brain and brain functions. Having in mind that it is an algorithm which can perform a specific task, we can use its capabilities to fight pandemics in several ways:
- To facilitate crisis response through an automated problem-solving process. This process would involve two types of tasks: a) discover tasks – natural grouping, anomalies, co-occurrence patterns, etc., and b) predictive tasks – predictions on the causes of the pandemics, recommendations, classifications, etc.
- Use the brain functions to get information as to how people are reacting during the pandemics based on sentiment analysis, see how the virus is spreading, and use a combination of the human capabilities and AI to improve the pace and accuracy of crisis analysis.
To sum up, we are living in times where we underestimate the power of data, AI, and technology to fight things that do not have much to do with technology but more with nature. This is our biggest mistake. We can employ data and AI everywhere, we just need to alternate the approach, and think out-of-the-box. If we had reacted faster and used our insights promptly, maybe COVID-19 wouldn’t have spread so much. However, this is a lesson that will teach us for the future.