Featured Product
This Week in Quality Digest Live
Innovation Features
Jennifer Lauren Lee
Solved at last
Fred Miller
University of Arkansas Division of Agriculture leads USDA-NIFA research partnership
Zach Winn
Autonomous machines snip weeds and preserve crops without herbicides
Daniel Croft
Noncontact scanning for safer, faster, more accurate, and cost-effective inspections
National Physical Laboratory
Using Raman spectroscopy for graphene and related 2D materials

More Features

Innovation News
Multimaterial 3D printer is integrated in the digital CAx process chain of Siemens NX
Strengthens data analysis and AI capability
Even nonprogrammers can implement complex test systems and standards
Alliance will help processors in the US, Canada, and Mexico
New features revolutionize metrology and inspection processes with nondimensional AI inspection
Strategic partnership expands industrial machining and repair capabilities
Supports robots from 14 leading manufacturers
AI designed to improve productivity and processes

More News

Ryan E. Day

Innovation

AI in the Sky: Deep Learning Drones

What will you teach your drone to do?

Published: Monday, October 11, 2021 - 12:03

Aquiline Drones Corp. (AD) quips that, “All roads lead to AI.” Recent developments in artificial intelligence (AI) integration make that statement hard to argue against, and the astute application of AI to AD’s cloud-based services also makes a lot of sense. So, smart drones are a real thing. But are they smart enough to solve problems and create a return on investment?

I’ve already written about AI in relation to the supply chain and automation, and with good reason. According to a report by Grand View Research, worldwide revenues for the AI market, including software, hardware, and services, are forecast to grow 40.2 percent annually, topping $997.77 billion by the end of 2028. Now, with AD acquiring ElluminAI Labs, I’m exploring AI in the realm of commercial aerial drones.

AI enabled drones

According to AD, the acquisition of ElluminAI Labs is the second strategic acquisition in the company’s pre-IPO plan. The first being the purchase of 50 percent of Netherlands-based Aerialtronics, a renowned drone manufacturer, for $9 million from Paris-based Drone Volt. The purchase was principally to bolster AD’s in-house R&D capabilities.

The acquisition of ElluminAI Labs enables AD to incorporate deep learning AI frameworks throughout its entire drone technology ecosystem which includes manufacturing, pilot training, drone business operations, and cloud solutions, all on a global scale.

AD’s proprietary AI framework is called Spartacus. Spartacus is modular with various sets of services and proficiencies. It can learn specialized skills on a case-by-case basis, while offering broad integration ability across various environments.

“Spartacus employs different types of models for its artificial intelligence,” says Barry Alexander, founder and chairman of Aquiline Drones. “These models can be trained for specific scenarios. As Spartacus gains experience, its accuracy increases. Each skill is modular. For example, a Spartacus model that is proficient at fruit counting would be employed in smart farming. Another model that is able to recognize thermal scans of fires to accurately predict hot spots can be used as a firefighting assistant.”

Barry Alexander CEO Aquiline Drones
Barry Alexander, founder and chairman of Aquiline Drones

Cloud-based operations

AI is foundational for AD to deliver key capabilities across the spectrum of their drone and cloud-based services. Not the least of which is a natural language processing engine that enables voice recognition and voice command and acts as an AI assistant across AD’s drone technology ecosystem. Natural language processing will play a role in AD’s pilot training dubbed, “Flight to the Future.” The assistant will answer questions and intervene with precise suggestions based on how students interact with the course curriculum.

“Spartacus is trained on educational content on a course-by-course basis and appears as a journey assistant to Flight to the Future participants,” explains Alexander. “It is available as a universal chatbot, including voice command and recognition, where students can ask questions. It also intervenes after a student has taken a lesson quiz. The intervention is based on the individual’s results. Because it is trained in the course content, the AI is able to effectively answer all sorts of queries about the course and curriculum.”

Spartacus is also equipped with machine vision skills to analyze drone data in real time. The data analysis is highly specialized by use case and Spartacus can master each application. For example, AD is currently working with a prominent commercial orchard to release specialized AI packages for data analysis, which include applications as diverse as fruit counting and fruit analysis. Other AI applications include the deciphering of thermal images from firefighting drones.

“Spartacus plays a crucial role in AD’s industry-leading robotic solutions for inspection, public safety, and future unmanned aerial vehicle (UAV) delivery services,” says Alexander. “For example, Spartacus will be part of AD’s command and control framework, optimizing flight planning and execution. It is also used for dynamic scheduling and flight dispatching of fully autonomous solutions.”

Deep-learning apps

Alexander maintains that the main difference between Spartacus and other digital agents is its advanced architecture, based on artificial neural networks, which provide the ability to get smarter and more efficient over time. This is notable in the current climate of ever-increasing amounts of data.

“Our acquisition of ElluminAI will produce a powerful system that delivers a modular, deep learning cognitive framework, capable of flight optimization, training assistance and data analysis for the most sophisticated business processes involving commercial drones to essentially improve a company’s profitability,” claims Alexander. "Most significantly, the deep learning framework behind Spartacus allows for big data analysis. As experience is gained, Spartacus has the potential to master multiple knowledge spheres discerning hidden patterns within the aggregated data. This has significant implications for future machine operations and positions the AD cloud as an industry transformer for commercial drone operations.”

Opportunities for ROI

The value of deep learning as a component of ROI is significant.

“Unlike traditional machine learning, Spartacus’ deep learning doesn’t reach a point of diminishing returns as data gets more complex,” adds Alexander. “In fact, Spartacus becomes smarter with experience as individual skills are created for different use cases and conditions.”

Alexander highlights the role of Spartacus in the handling of UAV traffic across the National Airspace System through AD’s command and control capability on the AD cloud. Here, Spartacus gathers data from the Federal Aviation Administration and other official sources, combines it with weather and additional external information, to provide crucial, real-time notifications and advisories to operation planners and drone pilots before and during the mission.

“Spartacus can be enhanced for different internet of things (IoT) settings, such as factory management, logistics, construction-site surveillance, precision farming, facility management, and maintenance advisor—to name a few,” boasts Alexander. “Each role requires building the right set of skills beforehand. With both drones and IoT devices becoming increasingly popular in such environments, Spartacus brings capabilities to the table that just cannot be ignored. Spartacus will play a significant role in UAV traffic and mission management. In AD's command and control, advanced flight control features will utilize both artificial neural networks and support vector regression to optimize flight ops. For completely autonomous missions, Spartacus can also assist in optimizing the scheduling and dispatching of flights, based on real-world conditions and equipment health.”

Discuss

About The Author

Ryan E. Day’s picture

Ryan E. Day

Ryan E. Day is Quality Digest’s project manager and senior editor for solution-based reporting, which brings together those seeking business improvement solutions and solution providers. Day has spent the last decade researching and interviewing top business leaders and continuous improvement experts at companies like Sakor, Ford, Merchandize Liquidators, Olympus, 3D Systems, Hexagon, Intertek, InfinityQS, Johnson Controls, FARO, and Eckel Industries. Most of his reporting is done with the help of his 20 lb tabby cat at his side.