MBA students combine AI and drones, turning project into product

A group of five Professional Flex students created a high-tech tool with the potential to change an industry.

Sometimes, with the right people, the right skills, and a bit of persistence and imagination, a class project can turn into reality.

A group of Professional Flex MBA students found this magical combination and came up with a high-tech tool with the potential to change an industry.

The team of five come from different backgrounds and industries. Archana Pailla has a background in computer science and engineering and is a senior software engineer at PayPal. Amanda Williams, who has a marketing and finance background, works at Vanguard. Prashant Bolla is an electrical engineer at Intel. Kristen Renee has a financial background and works at OneAZ Credit Union corporate headquarters in Phoenix. John Weyenberg also is in finance and is a commercial underwriter at Alliance Bank of Arizona.

They met in CIS 503. Professor Collin Sellman tasked the class to develop a cutting-edge business idea based on data and advanced analytics.

“Their project was a great example of applying artificial intelligence (AI) and robotics to a real problem in the construction industry, so they truly captured the essence of the assignment,” Sellman says.

The team used their unique backgrounds and experience to create a drone that could autonomously fly over and through construction sites and survey projects. The drone’s software would then compare the data it’s collected to the blueprints of the finished project to provide an estimate of completion. The project name: Phone-A-Drone.

The rise of Phone-A-Drone

The idea stemmed from Weyenberg’s work. He works with commercial lenders and construction developers. For funds to be released for a project, the bank needs to know how much more work is needed to complete the project. The bank would send an inspector out to survey the project and then report back on their findings. The whole process takes a considerable amount of time, Weyenberg explains.

The team wondered if the process could be streamlined using technology instead of people. “We thought we could incorporate technology into that plan and start using drones for monitoring the construction and estimating how much of it is complete through data analysis,” Bolla says.

The drone wouldn’t only take pictures and videos. It would also use GPS to find the construction site. Then the drone’s software would survey the exterior and interior using cameras and advanced technology like Light Imaging Detection and Ranging (LIDAR), and lastly, transmit data back to a hub.

What makes Phone-A-Drone unique is that the team would use machine learning to teach the drone to compare the original blueprints of the construction project to what it’s seeing on-site. The team would upload project plans into the drone’s software. Then the drone would analyze the data it collected and be able to send back an estimate of construction completeness. Based on that analysis, bankers would be able to estimate how much funding the project requires.

The drone can’t analyze on its own yet. The team would need to spend time inputting information or training the drone’s artificial intelligence. Pailla estimated that it would take two to three years of feeding information to the drone to train it. “That’s the beauty of machine learning and artificial intelligence,” Pailla says. “Any machine can do some things by itself, but for it to become more intelligent, you need to give it training and feedback.”

Relying on machines to do the work, instead of people, could streamline many processes that save time and money for the construction developer and lender, Renee says.

Using drones in the construction industry is an idea that’s taking off. According to a research study by Skylogic Research, “Goldman Sachs estimates that of the total global spending over the next five years on drones in the commercial market, about $11.2 billion will be generated by the construction industry, with $1.3 billion in the U.S. alone. The team estimates that by pairing drone technology with sophisticated models, Phone-A-Drone could capture a significant portion of the market potential.

There is competition in this corner of the market. Some drone surveillance companies offer basic fly-over photo services — even real estate companies offer a drone’s eye view of a house in the listing. But Phone-A-Drone’s concept would go a step above and offer superior data analytics technology. The drone would use artificial intelligence to analyze and compare what it’s seeing and then provide feedback — a feature competitors aren’t doing yet.

“That’s the biggest strength of our project, that it all can be automated,” Pailla says.

The team built a prototype of the drone. The expense of the hardware — the drone itself — wasn’t too much, Bolla says. The real cost to commercialize their idea would come from the software to train and control the drone. For the technology to recognize, for example, a house that’s half-built, it would require 1,000 pictures of the house and some images that aren’t of the house for the drone to learn to recognize the difference, Bolla says.

While the team will have to consider hurdles such as FAA and city flight regulations, they envision this technology being used for commercial developments and some large-scale residential subdivisions. Bolla says the team has thought about harvesting the technology to use in other sectors such as mining and agriculture.

Team members apply skills to project roles

Each team member had a unique role in the project. Pailla’s focus was on machine learning and data science. She worked on the drone’s software and focused on teaching it to analyze data and think for itself. Bolla worked on the hardware for the project and developed a prototype drone. Williams focused on marketing and the team side of the project, and Renee and Weyenberg used their financial backgrounds to connect with the customer and user side of the product.

The group earned an A on the project.

Professor Sellman says they had a great paper, demoing the drone during their class presentation.

“I am quite confident that ideas like this will reinvent a variety of industries as technology continues to advance,” Sellman says. “Drones are already being used in agriculture and several other industries, so as AI becomes more advanced, it is only a matter of time until autonomous drones are applied to real problems like the ones in this project.”

Class project turned product

Turning their class project into a product to take to market is something they’ve all considered and would like to make a reality.

The problem standing in their way is time. As working professionals and students in a graduate program, they’ve moved on to other classes and other class projects and don’t have the time necessary to devote to Phone-A-Drone. They plan to come back together after graduation in May and see what they can do to turn their idea into an actual business.

In the meantime, the team will continue to work together in their second year of the MBA program. Not only do they work well together, but they’ve also become friends.

“All our backgrounds and personalities balance each other out,” Williams says.