Unlocking the real value of drone technology

Debate has been swelling up within drone industry circles on the place of human pilots along the drone value chain. Informal conversations with drone services providers have indeed revealed the common theme that the drone industry is fast gravitating towards more autonomous operations.

Yea. We know this is not something drone pilots want to hear, having invested time and money in a skill that – as the below article explains – is making drone adoption so costly for organisations that they are having to consider alternative options where drone can easily be more effective with the final output.

The bottom line is; we know drones are better. But how do we make them scalable? That is the question American Robotics Cofounder and CEO Reese Mozer attempts to address in the below article, which first appeared in Venturebeat.

It does make sense of why there seems to be a rush for drone nests and autonomous drone innovation among drone makers of late.

We hope you like it.

Drones have been talked about extensively for two decades now.

In many respects, that attention has been warranted. Military drones have changed the way we fight wars. Consumer drones have changed the way we film the world.

For the commercial market, however, drones have largely been a false start. In 2013, the Association for Unmanned Vehicle Systems International (AUVSI) predicted an $82 billion market by 2025. In 2016, PwC predicted $127 billion within the “near future.” But we aren’t anywhere close to those projections yet.

Why is that? 

The evolving impact of graph analytics

Let’s start with the primary purpose of drones in a commercial setting: data collection and analysis. The drone itself is a means to an end – a flying camera from which to get a unique aerial perspective of assets for inspection and analysis, be it a pipeline, gravel storage yard, or vineyard.

As a result, drones in this context fall under the umbrella of “remote sensing.” 

In the world of remote sensing, drones are not the only player. There are high-orbit satellites, low-orbit satellites, airplanes, helicopters and hot air balloons.

So, what is it that drones have that the other remote sensing methods do not?

The first simple answer: image resolution

What “high resolution” really means

… for one product’s high resolutions is another’s low resolution…

Image resolution, or more aptly Ground Sample Distance (GSD) in this case, is a product of two primary factors: (1) how powerful your imaging sensor is, and (2) how close you are to the object you are imaging.Because drones are typically flying very low to the ground (50-400 feet above ground level), the opportunity to collect higher image resolutions than aircraft or satellites operating at higher altitudes is significant. Eventually you run into issues with physics, optics and economics, and the only way to get a better picture is to get closer to the object.

To quantify this: 

  • “High resolution” for a drone operating at 50ft AGL with a 60MP camera is around 1mm/pixel. 
  • “High resolution” for a manned aircraft service, like the now-defunct Terravion, was 10cm/pixel. 
  • “High resolution” for a low-orbit satellite service, like Planet Labs, is 50cm/pixel. 

Put in another way, drones can provide upwards of 500 times the image resolution of the best satellite solutions. 

The power of high resolution

Why does this matter? It turns out there is a very direct and powerful correlation between image resolution and potential value. As the computing phrase goes: “garbage in, garbage out.” The quality and breadth of machine vision-based analytics opportunities are exponentially higher at the resolutions a drone can provide versus other methods.

A satellite might be able to tell you how many well pads are in Texas, but a drone can tell you exactly where and how the equipment on those pads is leaking. A manned aircraft might be able to tell you what part of your cornfield is stressed, but a drone can tell you what pest or disease is causing it. In other words, if you want to resolve a crack, bug, weed, leak or similarly small anomaly, you need the proper image resolution to do so.

Bringing artificial intelligence into the equation

Once that proper image resolution is obtained, now we can begin training neural networks (NNs) and other machine learning (ML) algorithms to learn about these anomalies, detect them, alert for them and potentially even predict them.

Now, software can learn how to differentiate between an oil spill and a shadow; can precisely calculate the volume of a stockpile; or measure a slight skew in a rail track that could cause a derailment. 

American Robotics estimates that over 10 million industrial asset sites worldwide have found good use for automated drone-in-a-box (DIB) systems, collecting and analysing over 20GB of data per day per drone. In the United States alone, there are over 900,000 oil and gas well pads, 500,000 miles of pipeline, 60,000 electrical substations, and 140,000 miles of rail track, all of which require constant monitoring to ensure safety and productivity.

As a result, the scale of this opportunity is actually hard to quantify. What does it mean to fully digitize the world’s physical assets every day, across all critical industries? What does it mean if we can start applying modern AI to petabytes of ultra-high-resolution data that has never existed before? What efficiencies are unlocked if you can detect every leak, crack and area of damage in near-real time? Whatever the answer, I’d wager the $82B and $127B numbers estimated by AUVSI and PwC are actually low.

But now that we have established that the opportunities for drone technology are so large and clear, why haven’t these market predictions come true yet?

Enter the second important capability unlocked by autonomy: imaging frequency.

Autonomous drone systems proponents, American Robotics and Airobotics were recently brought under the aegis of the Ondas group

What does “high frequency” really mean?

The useful imaging frequency rate is 10x or more than what people originally thought.

The biggest performance difference between autonomous drone systems and piloted ones is the frequency of data capture, processing and analysis. For 90 percent of commercial drone use cases, a drone must fly repetitively and continuously over the same plot of land, day after day, year after year, to have value.

This is the case for agricultural fields, oil pipelines, solar panel farms, nuclear power plants, perimeter security, mines, railyards and stockpile yards. When examining the full operation loop from setup to processed, analysed data, it is clear that operating a drone manually is much more than a full-time job.

And at an average of $150/hour per drone operator, it is clear a full-time operational burden across all assets is simply not feasible for most customers, use cases and markets. 

This is the central reason why all the predictions about the commercial drone industry have, thus far, been delayed. Imaging an asset with a drone once or twice a year has little to no value in most use cases. For one reason or another, this frequency requirement was overlooked, and until recently (subscription required), autonomous operations that would enable high-frequency drone inspections were prohibited by most federal governments around the world. 

With a fully-automated drone-in-a-box system, on-the-ground humans (both pilots and observers) have been removed from the equation, and the economics have completely changed as a result. DIB technology allows for constant operation, multiple times per day, at less than a tenth of the cost of a manually operated drone service.

With this increased frequency comes not only cost savings but, more importantly, the ability to track problems when and where they occur and properly train AI models to do so autonomously. Since you don’t know when and where a methane leak or rail tie crack will occur, the only option is to scan every asset as frequently as possible. And if you are gathering that much data, you better build some software to help filter out the key information to end users.

Tying this to real-world applications today

Autonomous drone technology represents a revolutionary ability to digitise and analyse the physical world, improving the efficiency and sustainability of our world’s critical infrastructure.

And thankfully, we have finally moved out of the theoretical and into the operational. After 20 long years of riding drones up and down the Gartner Hype Cycle, the “plateau of productivity” is cresting.

In January 2021, American Robotics became the first company approved by the FAA to operate a drone system beyond visual line-of-sight (BVLOS) with no humans on the ground, a seminal milestone unlocking the first truly autonomous operations. In May 2022, this approval was expanded to include ten total sites across eight US states, signalling a clear path to national scale. 

More importantly, AI software now has a practical mechanism to flourish and grow. Companies like Stockpile Reports are using automated drone technology for daily stockpile volumetrics and inventory monitoring. The Ardenna Rail-Inspector Software now has a path to scale across our nation’s rail infrastructure.

AI software companies like Dynam.AI have a new market for their technology and services. And customers like Chevron and ConocoPhillips are looking toward a near-future where methane emissions and oil leaks are significantly curtailed using daily inspections from autonomous drone systems

My recommendation: Look not to the smartphone, but to the oil fieldsrail yardsstockpile yards, and farms for the next data and AI revolution. It may not have the same pomp and circumstance as the “metaverse,” but the industrial metaverse might just be more impactful. 


Leave a Comment


Welcome! Login in to your account

Remember me Lost your password?

Lost Password