How advanced analytics pulls insights from the weather, part 1
Natural disasters seem inescapable, leaving us feeling vulnerable in the hands of nature. How is this possible, given all the data and technology that surrounds us today? Can’t experts get better at prediction, and even try to stave off more natural calamities or more effectively reduce the loss of resources?
The answer is both yes and no. Sometimes we know a tornado will hit and we cannot prevent the loss from that situation. We know floods will happen due to hurricanes, and locals can strive to effectively reduce the damages. But they cannot usually prevent damage from happening. We know ice storms and hail storms will likely occur, but we cannot always prevent damages to homes or cars.
While it can be difficult to predict the forces of nature, we can use machine intelligence to effectively capture more useful data from these situations and learn better. Experts can apply machine intelligence and help future generations become better at managing resources more effectively and mitigating disaster.
One of the finest examples that I see today is the application of IBM Watson cognitive computing to predict weather, including formation of clouds, rain, temperature and more. IBM has actually been working on weather for years. Since the 1990s, IBM has collaborated with the U.S. National Oceanic and Atmospheric Administration (NOAA) to help the government with short-term forecasting models. The company has long been a source of validation for both inner-earth and outer-space weather.
How does cognitive computing help us predict the weather? You can apply machine intelligence—or in simple terms, making machines represent human thinking and reaction—without emotion. This is the apex of artificial intelligence and automation applied together. It’s the balance of intelligence across the processing spectrum and how to make decisions is where the build of this ecosystem lies.
What industries or organizations that actually consume weather-related data intelligence? What are they doing with it? Let’s take a look at several scenarios:
- Farmers are one of most weather-focused groups of professionals across the world. Advanced analytics can help farmers analyze real time data like weather, temperature, and moisture. Data can lead to insights on how to optimize and increase yield, improve farm planning, make smarter decisions about the level of resources needed, like when and where to distribute them in order to prevent waste.
- Airlines are another major consumer of weather data. It helps them plan and optimize aircraft traffic, plan arrivals and departures, plan critical activities like de-icing, provide additional cooling in extreme heat and much more. The ability to introduce machine learning and cognitive computing makes a lot the work here become more insightful and help drive efficiency.
- NASA has a considerable amount of equipment in space that ingest data from weather systems. The agency uses the data to understand the effect of how, for example, a solar flare or an asteroid could affect communications between space equipment and the space station. NASA’s use of algorithms has created efficient systems that are used more widely in industry and academic contexts. One example is the Mars Rover project, which we will discuss in greater detail in an upcoming blog post.
- Real estate companies are turning to analytics to make buildings more energy efficient. Many modern buildings now feature a myriad range of sensors that track heat, light and more. But re the systems inside the building responding based on sensor data? Could machine learning add efficiency? In the future, are there neural network layers where builders can find even more value for residents or workers?
- Finally, the energy industry is another huge driver of change for the machine learning adoption and integration of weather analytics. Some of today’s energy producers are generating power using windmills and solar panels. The sensors at these windmills and solar panel stations are generating a lot of data, and the data can be harnessed to provide better data-driven decision-making. Imagine what’s possible when engineers add machine learning and artificial intelligence to help optimize outcomes.
In my next blog post, we’ll examine the other side of the analytics ecosystem and look at the scenarios from the side of the organizations producing all the data consumed by the above industries.
Interested in learning more about IBM analytics, and how industries are making better data-driven decisions on the road to AI? Read more in our interactive guide to hybrid data management.
This article is offered for publication in association with the Big Data Seminar 2018 NYCity, October 25-26, 2018 in New York City at Hotel Pennsylvania, 401 7th Avenue, New York, N.Y., sponsored by Data Management Forum. Additional information is available in the Big Data Seminar flyer . For additional information please call (516) 221-5560 or email here.