DART

Meteorology has a massive impact on each of our lives. From the app we check for the temperature every morning to the prediction of major weather events and natural disasters, meteorology is a tool that many of us couldn’t, or at least wouldn’t want to, live without. While modern meteorology has certainly improved over the last several years, there is one thing most technologies in this field cannot employ: real-time, micro-level precision. 

Sujit Sinha takes a different approach. Sinha is developing precision meteorological prediction employing a Data-Driven, Adaptive, Real-Time (DART) approach. This technique is being utilized to improve highly-localized meteorological forecast accuracy. The technology works by using data collected from instruments onboard uncrewed aerial vehicles (UAVs) to drive the adaptation of a weather simulation model in real time. 

“This technology has the ability to improve weather forecasting on a microscale within a 50-meter by 50-meter area, which is a very small, almost house-by-house region, allowing applications from agriculture to military, air transport, and more,” said Sinha. 

In order to model weather with such precision, the DART approach uses a retrospective cost adaptation (RCA) algorithm that is often used for system control applications, such as missile guidance.

Before beginning his journey with this technology, Sinha built quite an impressive career, working as a flight controller on the Space Shuttle program for National Aeronautics and Space Administration (NASA), American Airlines, Department of Defense at the Pentagon, a management consultant to CEOs at aerospace and high-tech firms, and as an executive at Motorola at different points in his career. Before his retirement, Sinha even dabbled in the apparel space, serving as Chief Information Officer for a company that produced dance clothing.  

Sinha earned his bachelor's degree in mechanical engineering from the University of Kentucky before going to University of Alabama in Huntsville and then The Wharton School of the University of Pennsylvania, where he earned a master’s in engineering and a Master of Business Administration, respectively. 

After retiring, Sinha hoped to find a meaningful project that would take him back to his roots: aerospace engineering. This led him to pursue a Ph.D. in aerospace engineering and join a team of five professors and several graduate students who were working on the DART approach, a project that he ultimately learned about as a member of the UK College of Engineering Alumni Board. 

To set the technology and accompanying RCA algorithm into motion, the team uses drones. These drones fly in a specific formation and gather data to feed information into a weather model. Then, the revised weather forecast is fed-back into the guidance system of the drones. From there, this technology will reposition the drones to obtain the next most beneficial set of data. Aside from forecasting on an incredibly small scale, this feedback loop into the drone guidance system is a key aspect that makes this operational concept unique compared to other weather model technology. 

“The most prominent industries that this technology could benefit include agriculture and emergency services/public safety,” said Sinha. This tech has the potential to aid farmers in predicting when crops should be harvested, far sooner in season than this decision can typically be made. This technology can also track contaminant cloud leaks, which are created during chemical or other disasters. Effectively and precisely tracking these leaks in the way that would aid first responders in knowing exactly how they should evacuate people in its path. Outside of these uses, improved identification of highly-local, hazardous weather phenomena would also benefit the military, emerging drone package delivery firms, such as Amazon Prime Air and Alphabet’s Wing, and future air taxi services.

Sinha says there are four main development tasks of this operation: instrumentation, autonomous flight guidance and control, the weather simulation model coupled with the use of the RCA algorithm, and testing. He let us know that the development team is making progress in all these areas. The instrumentation and automated flight aspects are coming along, and the team has also generated the first true flight test data. In addition to these milestones, the team recently filed a provisional patent and got the weather simulation model working with the RCA algorithm. 

Sinha said that Launch Blue’s impact on understanding UAV commercial and military operations, as well as the precision meteorology market, has been immense.  “It forced me to go talk to interested communities and stakeholders,” Sinha said.

Through his participation with Launch Blue, Sinha was able to network and build connections with “venture capitalists, people in forestry who use drones for fire monitoring, military, commercial operators, and even industry and non-profit organizations.” Sinha’s experience with Launch Blue allowed him to meet some of his target audience, as he was introduced to those who expressed a need for the type of information that the DART solution can provide.  

The team plans to further develop the technology and continue talking to those in various industries where improved precision meteorological forecasts could play a role. It's just enough to keep Sinha busy while working to earn his doctorate.  

By: MaKenzie Purdom

Launch Blue nurtures promising startup founders and university innovators through intensive accelerator and incubator programs. Its funding partners are the University of Kentucky: Office of Technology Commercialization, KY Innovation, the U.S. Economic Development Administration, and the National Science Foundation.