IBM Big Data & Analytics Hub

Many of us seem to watch weather forecasts to figure out what to wear the next day but forget about it right after that, unless of course there is snow in the forecast. Especially here in the northeast; we dread watching the weather report for about six months of the year.

Image Courtesy: Gary Varvel

For this reason, IBM’s acquisition of The Weather Company was a head-scratching moment for many because we are used to only the consumer aspect of weather, not the business side—especially given the high speculation by The Wall Street Journal.

Why did IBM, an IT software company, go after The Weather Company then? IBM started this fundamental shift a few years ago, transforming itself from a big IT and mainframe provider to a digital, data and insight company. Recent speeches by the CEO at IBM clearly articulate its main focus has shifted toward cognitive computing, analytics, Internet of Things, application programming interfaces (APIs), hybrid cloud and digital platforms that support big corporations to reinvent themselves and engage in the digital economy.

Acquire the ability to make intelligent predictions

With the frenzy to digitize everything, just compute power or a platform is not enough for big corporations. There are multiple cloud providers that can provide those commoditized services cost-effectively. Instead, values for business come from providing data, and, more importantly, predictable results along with actionable intelligence. Hence, this investment and others to secure new data sources—such as Merge Healthcare ($700 million) along with other new relationships including Apple, Johnson & Johnson, Medtronic and Twitter—align well with IBM’s strategy.

Consider how weather affects business. Routine weather cost US businesses over $500 billion in 2014. Outside of political turmoil, weather is the most powerful force that can interrupt a business. Predictable weather information—using the three billion global forecast reference points—and its effects on business in the near or far future, and actionable insights to avoid such situations, could be worth more to companies than just a weather forecast.

For example, consider the airlines and weather. We all know how much we hate to be in a plane when it hits turbulence. What we don’t know is that the weather patterns affect airlines in many more ways than just hitting turbulence. Here are some scenarios in which weather prediction as a service is helping airlines to save money today.

Scenario 1: Airport congestion delays

Airlines tend to carry just enough fuel to avoid extra fuel burn. An average airline with 1,200 average daily flights can save as much as $1.6 million of fuel burn by not carrying extra fuel for airborne-holding reserves. This figure is not even taking into account the greenhouse effect and the environmental issues created by the extra fuel burn.

Unless you are able to predict the weather and both the arrival and departure pattern at the same time, you can’t predict the average taxi time and actual departure time of a plane. For an average airline, that time could be costly. For example, an average airline with the same rate of average daily flights can receive an annual benefit of $534,000 by avoiding carrying excess taxi-out fuel. But, by avoiding under-carrying taxi-out fuel, they can also save $247,000. As you can see, this example represents a double-edged sword and needs to be balanced out.

This scenario is an actual use case from The Weather Company predictions at John F. Kennedy International airport. On 20 August 2015, a major thunderstorm moved through the area, and the departure traffic was going to be delayed as a result. At about 21:00 hours, the airport congestion analytics modeling software predicted that the departure congestion would last for another five hours. The taxi-time prediction software predicted that the time to taxi out could be as high as 50 minutes for the first three hours. Based on this insight, dispatchers added contingency fuel for departing flights, reducing the risk of costly fuel-related diversion or an unnecessary stop to fill fuel before takeoff. That reaction enabled just enough extra fuel for the planes to account for extra taxi time, and they didn’t have to carry that fuel in the air or have to pull over to fuel up right before takeoff.

Scenario 2: Turbulence subjectivity

Another interesting use case from The Weather Company has to do with turbulence during flights. Many major airlines struggle with its affects today in several ways. These ways include accurately predicting flight paths, dynamically rerouting planes efficiently in the safest possible way and, most importantly, pulling aircraft out for service to comply with Federal Aviation Administration (FAA) standards when planes hit certain levels of turbulence.

But a lot of subjectivity is involved in these actions. For example, a certain group of overconfident pilots may report the turbulence as normal, so the airline may miss the opportunity to service the plane on time. Another pilot group may be less confident and report every bump they face as major turbulence, so the planes are hauled out of service for unwarranted maintenance that can cost the airlines millions of wasted dollars because of unnecessary downtime.

Combine this situation with the FAA’s strict compliance regulations and the magnitude of the problem only increases. According to the US National Transportation Safety Board (NTSB), turbulence accounted for 66 percent of FAA Part 121 weather-related citations. Of those citations, 93 percent resulted in injuries that occurred when the majority of aircraft had the seat belt sign illuminated. By choosing a path to avoid that turbulence, the incident, the citation and the unnecessary maintenance downtime can be easily avoided.

To avoid these situations, a major airline has fitted most of its planes with onboard Internet of Things devices to measure turbulence activity. These devices accurately record the duration and location of turbulence levels. They also send a message to ground computers immediately when a threshold is exceeded. The data collected is assimilated into forecasts of flight-route hazards and automatically updates the flight paths of flights in the air or flights that are about to depart. Onboard devices also send weather alerts back to the plane for the pilots’ attention if a manual intervention is required. The airline also uses this objective information to decide which planes need to be serviced as opposed to relying purely on the subjective log information provided by pilots.

In the picture above, you can see the flight path for a Kalitta Air flight (white line) from Anchorage, AK to Hong Kong, was dynamically altered based on accurate measurements and prediction from The Weather Company using its WSI Fusion product to avoid icing, convection and turbulence. (Light blue circles are icing areas).

According to Bryson Koehler, the CIO & CTO for The Weather Company, “The Weather Company has built a massively scalable IoT Data Platform they call SUN (Storage Utility Network) to power upwards of 26 billion requests per day and leveraging this Data Platform to become a foundational service for all of IBM to use for solving new types of problems driven by weather and/or IoT data sets is what this deal is really all about.”

Take a transformative course of action

When placing bets for your enterprise digital transformation, you need an organization that can provide a transformative solution to take you into the future, not just a platform that can barely solve your current problems. Continue this conversation with @AndyThurai to learn more about how IBM can help with your company’s digital transformation.

This article was originally published in IBMbigdatahub on nov 9, 2015 – https://www.ibmbigdatahub.com/blog/business-value-weather-data

–