Weather Decision Technologies
Actionable Weather Information to Your Rescue

As a menacing tornado churned behind him, sensing the only way out to be speeding directly at it, the storm chaser did just that, and floored his truck to safety. Weather Decision Technologies’ President and CEO Mike Eilts remember this event from more than two decades ago like it was only yesterday. He is adamant observations of atmospheric conditions such as wind, temperature, pressure and humidity collected while tracking storms are vital for scientists to understand dangerous weather. The data delivered as a result is essential in developing increasingly accurate warning systems that provide the public ample lead time to seek shelter. The passion to incorporate this data led Eilts and four others, all from similar backgrounds in meteorological research organizations in Norman, OK, into founding Weather Decision Technologies (WDT) in 2000. The goal–to utilize specific expertise with data analytics and large volumes of real-time data as it applies to hazardous weather detection and prediction, forecast modeling, decision analytics, GIS, mobile apps and interactive mapping.

WDT uses its analytics engines, weather models and a decision engine to provide customized alerts, now casts and forecasts

Curious about how the 2014 severe weather season might play out? Eilts explains—“The biggest challenge for the weather industry when using Big Data solutions is not in the volume of the data we deal with, but instead in the fact weather data is very volatile. This means every few seconds we get new data and every hour we get results from models we run internally.” WDT aggregates petabytes of weather data from global sources such as surface observing stations, radars, lightning networks, and satellites. The company’s proprietary capabilities include processes to control the quality of raw data.

WDT’s flagship product is a SaaS solution called WeatherOps® which is delivered through a web based portal and interactive map, along with a mobile app. When clients provide information about their asset locations and the specific weather phenomena that impacts them, the company uses its analytics engines, models and a decision engine to provide them with customized alerts, now casts (short-term predictions) and forecasts. To provide this type of geospatial decision support information to clients globally, WDT leverages Big Data storage and processing solutions to organize and manage data and produce analytics. For example, in abstract, the company handles mass amounts of data in real-time, distilling it down to short messages such as “lightning threat at a particular location from 9:15 – 9:30 p.m,” or alert messages for any given location globally on the duration of rainfall.

Weather is a very horizontal industry, meaning that high-value weather information that provides a customer with actionable information can be used by most industries. WDT focuses on a few key verticals including energy, aviation and local media. One of WDT’s longstanding clients is a large energy exploration and production company that has assets at over 100 locations across North America. In a recent severe thunderstorm event, WDT’s analytics and alerts provided this client with information to move personnel to safe locations before tornadoes caused damage at one of their facilities, while another location was threatened by lightning at the same time. Even during this time, they were able to keep production going by moving employees to a safer location until an all clear was provided by WDT.

WDT’s expert data scientists and software engineers are continuing to develop new methods to handle the combination of more dense and volatile data. The team is creating more complex analytics engines, pushing the envelope in computing resources, cloud services and Big Data solutions. WDT is on a strategic vision to continue to expand globally. “We are canvassing the globe to access new data sources like Doppler Radars and surface observations. We are also running our models at ever increasing resolutions and over larger and more diverse domains,” says Eilts.