Feb 10th 2017

A Toy Plane led to Our Current Autonomous Systems


Today, the drone industry is one of the most sophisticated in the world. But what exactly is a drone? It can be anything from a cheap remote-controlled toy to the U.S. Navy’s strike fighter-sized unmanned aircraft X-47B UCAS-D—which was designed and built by Northrop Grumman and uses a software-controlled smart “brain” to autonomously conduct aerial missions.

The X-47 Unmanned Combat Air System (Northrop Grumman)

Definitions vary on what can be considered a drone, but broadly speaking, it is a device that can take an action without human intervention. This differs from simple remote control, although the importance of remote operations particularly in the early days, cannot be overlooked. It should be noted that although “drone” is a popular moniker, it’s an oversimplification; it doesn’t adequately capture the fact that we are also dealing with autonomous aircraft with smart software brains.

Taking the above into account, what was the first “drone” and how did it evolve? Let’s take a deeper look.

The First Drones

Let’s take a look back to 1898, when Nikola Tesla used a radio-controlled motor to wow patrons at the Electrical Exhibition in Madison Square Garden by piloting a toy boat with “verbal commands,” while really manipulating a controller at the same time as his shouts.

Tesla’s radio-controlled boat (Nikola Tesla Museum)

This early application of radio controlled technology led to a small set of hobbyists interested in radio controlled devices. Just as those hobbyists do in their garages today, some early work led to very useful commercial devices.

In 1935 one hobbyist, renowned actor Reginald Denny, managed to show the U.S. Army that his hobby could be a useful target for antiaircraft gunnery crews. By the end of decade Denny had formalized the Radioplane company and was selling the first vehicle, the OQ-1, to the Army.

Since they were used primarily as targets to be shot down, thousands of the Radioplane vehicles were made in the 1940s. Some of the unassuming devices were even assembled by Marilyn Monroe, who was discovered by an army photographer sent to take pictures of the Radioplane in production.

By 1952, Northrop Aircraft purchased Denny’s company, which then became the Radioplane Division. This was to be the launch of Northrop Grumman’s development of autonomous systems. In the decades since, the systems and technologies have expanded vehicles to be used for much more than just target practice.


A desire to keep people away from dangerous conditions has long driven inventors to devise mechanisms that can go places and do things that people cannot. But the earliest drones — those that fit the definition of fully unmanned and automated — were unreliable by today’s standards.

The MGM-57 Falconer, designed in 1955, was the U.S. Army’s first unmanned reconnaissance aircraft, used for short-range battlefield surveillance. There was no way to transmit images in real time, so the onboard camera film had to be recovered and processed before use. Furthermore, the Falconer could only remain airborne for half an hour.

During the Vietnam War, the Ryan AQM-91 Firefly and Model 147B Lightning Bug were developed for long-range reconnaissance missions into China. Launched from other aircraft, these drones were self-guided and powered.

By the late 1970s, the Aquila, the most promising drone of the time, crashed every 20 flight hours and needed 30 people to launch it.

In 1977, Israel-based aerospace engineer Abraham Karem arrived in the United States. Working from his home in Los Angeles, using cheap materials, Karem set out to build a drone based on “inspired design and highly optimized and integrated subsystems.” In 1980, his drone, code-named the Albatross, stayed aloft for 56 hours. Karem thought of drones as reliable long-term vehicles that should have as long of a service life as their human-piloted counterparts.

Even up to the 1990s, operating crews needed be located near the vehicle to ensure radio signals made the connection. It was not until the turn of the new century that there were reports of vehicles being controlled from 7,000 miles away.

These advancements occurred in part because of unmanned systems with operating times measurable in years rather than minutes, and the development of satellite monitoring technology.

The Tech Stack of Modern UAVs

A flying object that can be remotely piloted from anywhere in the world requires a deep technology stack, the lack of which limited the evolution of drone systems until the 1980s.

With the advent of the first microprocessor in 1970, devices once the size of control systems shrunk drastically and powered the development of enabling technology for remote operations. Meanwhile, the global positioning system (GPS), a global navigation satellite system conceived in the 1960s, became fully operational in 1995 and allowed for operations around the globe.

You can now view the first armed UAV in the Smithsonian National Air and Space Museum, although the current U.S. fleet of more than 10,000 drones is employed primarily for intelligence gathering, surveillance and reconnaissance.

21st-Century Drones

The technology in the drone industry have come a long way from Tesla’s toy boat. Even the term “drone” has evolved to describe an array of autonomous aerial systems. Though a drone itself refers to the complete vehicle, each one employs sensors, controls and power systems in various stages of development. For instance, today’s Gray Eagle uses STARLite Tactical Radar, integrated to optimize operations while also providing target signatures that aid in intelligence gathering.

STARLite Tactical Radar (Northrop Grumman)

But radar and other systems present on nearly all unmanned systems have their limitations. Several factors including advances in airframes, power sources, and both electronic warfare and air defenses are constantly driving the need to innovate and upgrade systems. The trend towards open modular architectures allowing for better integration of manned and unmanned systems will further push technology’s future. Even Moore’s law is coming to an end and has helped drive advances in competing computational platforms.

Neural networks and AI are becoming more capable of analyzing radar, optical and environmental data to provide onboard identification and classification of observed threats. This enables many more actions to be taken without the need for human intervention.

Drone Technology Beyond Defense

Beyond threat removal, drone technology can also allow for aid to remote locations or perform routine tasks. As defense needs continue to drive technology advancements, existing technologies can be pushed to consumer markets.

Already the use of auto-pilot and GPS-guided “return to home” features have opened the market. Everything from real estate listings to the opening scenes of blockbuster films are now aided by unmanned devices that can be used in everyday society (though FAA regulations still limit who can commercially pilot a device).

Today, the development of self-driving cars and delivery drones are hot topics for logistics companies and retailers alike. Advances now happening in drone technology are so rapid that we’ll see more innovation within the next decade than in all those years since Tesla’s boat.

Check Out These Science Articles Too