October 5, 2022
An article about integrating highly autonomous drones and peripherals into your infrastructure inspections.
Drone technology is advancing by the minute. What is an autonomous drone, and how can autonomous drones be used for inspections? This article will help you understand the technology to step up your inspection game — the easy way.
What is an autonomous drone? It is a drone that is autonomous.
Okay, what does “autonomous” mean? Now that’s a better question to ask.
Autonomy, according to Merriam-Webster, is a state of “self-directing freedom”. In a similar fashion, an entity is autonomous if its decisions and actions are capable of being “undertaken or carried on without outside control”. (Merriam-Webster Dictionary, 2022)
Autonomous machines are not new — in fact, they have probably been in your home your entire life!
Toasters are autonomous in that they automatically heat the bread for a certain amount of time before popping it out (though the human still chooses the toasting level).
Thermostats are autonomous in that they automatically regulate a temperature to within a few degrees (though the human still chooses the target temperature).
Navigation apps are autonomous in that they find the optimal route between two locations (though the human chooses the destination, and actuates the vehicle controls).
That’s great; now what about autonomous drones?
As the concept applies to drones, autonomy implies a certain freedom to act without direct human control or intervention. Every drone ever built has exhibited some form of low-grade autonomous behavior: keeping a constant rotor RPM, maintaining a stable attitude or altitude, following waypoints, etc. If you are a drone pilot, you have been flying autonomous drones all along without knowing it!
But this article is not about the everyday consumer drone; it is about integrating highly autonomous drones and peripherals into your infrastructure inspections to improve your business. So let’s get started.
Here is a short list of reasons that countless businesses are incorporating highly autonomous systems into their operations.
Let’s unpack each of these.
Let’s face it: while humans are tremendous assets, they are also some of your greatest liabilities in the field. Every time a human sets foot at a job site or industrial facility, there is a non-zero chance of a serious injury. This chance is increased when the work environment includes ladders, platforms, slick surfaces, sharp edges, falling objects, toxic chemicals, high voltage,….. the list of hazards is long. Why would you send a human into harm’s way when an inexpensive machine is plenty capable of doing the job?
Sending a highly autonomous machine increases the safety margin even more. Humans can make catastrophic errors when flying in close proximity to obstacles or near the edges of the drone’s flight envelope. At best, these errors will damage the drone and its sensors. At worst, they can damage sensitive infrastructure or injure nearby personnel. The machine precision of highly autonomous drones overcomes human error to reduce the likelihood of a mishap to negligible levels.
General George S. Patton is credited with stating that a “good plan today is better than a perfect plan next week”. Adapting this to your business, let us restate: “a good decision today is better than a perfect decision tomorrow”. The best business decisions are backed by data, so the corollary is “good data today is better than perfect data tomorrow”.
But what if you could get “perfect” data today, instead of waiting for “good” data tomorrow?
Herein lies the power of highly autonomous drones. Drones can cover the same area in a day that would take a human ground team over a week to cover—and the data is just as good, if not better. Data processing occurs at—you guessed it—machine speed. If you need actionable information today, your best chance is with a highly autonomous drone.
Infrastructure owners generally do not consider inspections to be a “one and done” type of endeavor; inspections typically occur on recurring cycles. This is some of the real beauty of autonomous drones: they can fly the same mission over, and over, and over, and over…. at the touch of a button. There is no need to re-plan a mission or attempt to manually fly the same points. Autonomous drones will repeat the mission with (you guessed it) machine precision. This level of repeatability gives infrastructure owners the most accurate time-lapse perspective of their property at a pace never before experienced in the industry.
We have already mentioned “machine precision” a few times, but here is another angle to consider. Have you ever sought a professional’s expert opinion (physician, automotive mechanic, lawyer, etc.) and then sought a second opinion from somebody else? Were the results the same?
This phenomenon highlights the bias that every human exhibits. The same person might give a different opinion on different days depending on their mood, even with the best of intentions. Moreover, humans are error-prone and tend to make mistakes when they are tired, hungry, or emotionally distressed.
Machines, on the other hand, do not make mistakes: they collect inputs, conduct some processing, and generate an output. (For some interesting reading, search for “deterministic algorithm”.) Machines will give the same answer, every time. Any unexplained behavior is probably due to a faulty or unanticipated collection of inputs to the algorithm.
Autonomous software processing takes this one step further by conditioning and formatting information into a ready-to-deliver product, at the touch of a button. This reduces the human’s transcription error, copying numbers from one document to another or miscalculating a result. Autonomous software enhances product quality and trustworthiness.
While some autonomous systems are standalone, most modern applications are built with some level of cloud-basing or data centralization. This means that they have the benefit of learning from massive amounts of data.
For example: a human proofreader could read through this article and spot any misspelled words that they are familiar with, but they are unlikely to spot a mistake on a word they have never seen before. Fortunately, modern word processors maintain data on every person that has ever used it, so they have a massive library of the most commonly misspelled words or misused phrases. The spellchecker used for this article has the benefit of millions of users’ documents in its history to offer suggestions.
What does this mean for your autonomous drone and data processing software? While the electrical lineman may see an anomaly for the first time and not know what it means, the autonomous data processing application has probably seen thousands, or millions, of examples, and it knows exactly how to identify and correct the problem.
How many people would a survey team send to manually survey a large area, say 20 x 20 km? Four people? Eight people?
How many people would be required with an autonomous drone survey? One, maybe two if the pilot chooses to use a Visual Observer.
Most militaries see the benefits in “unmanning the front lines” and your business should see the same benefits. More humans at the site means more hazard liability and higher costs. Which leads us to…
How much did that survey team of 4-8 people cost? They each have a salary, medical insurance, life insurance, and workers compensation insurance that is costing you money whether you are the employer or the customer. They arrived in several vehicles with maintenance and insurance costs that are contributing overhead costs. They come with expensive equipment and personal protective equipment (PPE) that also contribute to overhead costs. And they make mistakes that increase your business liability and risk of a serious and expensive incident.
Autonomous drones mitigate all these factors. Fewer people and higher-quality work goes directly to your bottom line. If you are a Drone Service Provider to the Critical Infrastructure industry, your customers will be pleased to receive such high-quality results from such a small team with such a short turnaround time. If you are a Critical Infrastructure owner, you will wonder why you didn’t switch to drones years ago!
As more global capital is invested into the drone industry, technological advancements will continue to reduce the human workload. Even today, it is now possible to plan a flight mission, conduct the flight, and deliver a useful data set and report to your customer with just a few strokes on the keyboard, never leaving your office.
Many people see “autonomous” as a binary condition: either a system is autonomous, or it is not. Hopefully, you now understand that “autonomy” is a spectrum, and does not necessarily apply equally to each step in a task. For example: the human chooses the temperature, and the thermostat maintains that temperature.
The ultimate goal of autonomous system design is to give the right amount of control to both the human and the machine, allowing each entity to use its strengths while covering for the other’s weaknesses. This level of “human-machine teaming” provides a synergy far greater than either entity could accomplish alone. Here is a graphic from the academic world that exemplifies the concept.
In this graphic, both the human and the Artificial Intelligence (AI) have high levels of control, but not over the same tasks. Recall the navigation app from above: the app excels at searching for the optimal route, but it is not well-suited for controlling a steering wheel or monitoring roadway hazards. In this example, the AI has high levels of control in selecting the optimal route, while the driver has high levels of control to manipulate vehicle controls and veto the AI’s navigation recommendations.
How can you construct a similar autonomous drone solution for your drone inspection business? Let us start by identifying the tasks. For now, let us ignore the purely business processes of client intake, task order receipt, invoicing, and accounting, and instead focus on the pilot-related tasks of drone infrastructure inspection.
Most “autonomous drone” solutions address one, maybe two, of these items, but we can do better than that. Let us take a deeper dive into how you might assemble a full-stack solution for your drone inspection needs.
The Model-Based Systems Engineering (MBSE) discipline uses a tool called the “Use Case Diagram” to describe the interaction of multiple entities as they perform certain tasks. In this simplified scenario, the human pilot and the automated software application(s) are distinct entities. The following diagram is an example of a highly autonomous full-stack solution through all phases of mission planning, data acquisition, and data processing. Autonomous software does not replace human involvement; complements it. The AI conducts those tasks that machines are best-suited for, while enhancing the human’s ability to conduct those tasks that humans are best-suited for. Focus on how the human and the machine work together to achieve optimal results.
(Forgive the Stickman which represents Autonomous Software; this just happens to be UML symbology for “Actor”, whether human or otherwise.)
In reality, there are no full-stack autonomy solutions that cover the entire right half of the diagram, nor should there be. Individual software applications optimize their workflows for specific portions of the mission flow, which allows the drone inspection enterprise to tailor its toolkit to the desired task. A more complex version of this diagram would identify multiple automated software packages and their individual contributions to the mission flow.
Did you recently purchase a highly autonomous drone? Great, you probably addressed “Flight Control” and maybe part of “Mission Planning”. For the rest, you might have to do the old fashioned way—that is, until you design your own full-stack autonomous drone solution.
As the Drone Service Provider, NOW is the time to introduce your clients to the appropriate Autonomous Drone solution. Why? Because if you don’t, someone else will. Take the lead, get out in front of the inevitable, and offer these solutions, with your organization driving the solution, presenting or demonstrating the pros-and-cons.
These solutions, while yes, they do reduce manpower requirements, they do not entirely alleviate the need for drone operators to manage and operate the autonomous missions. You may as well be the one providing this as a part of your offering to the client.
~click into this article to learn more~ “Will Autonomous Drones Replace the Drone Pilot?“
Scopito is a data processing solution that addresses industrial and enterprise needs for large-scale infrastructure inspection. Its strengths lie in the bottom half of the Use Case Diagram, as it was designed to ease all processing tasks from data acquisition to product delivery.
Scopito is platform-agnostic, meaning that it can ingest data from any drone and any optical sensor to produce actionable information. You can upgrade your equipment and seamlessly continue using Scopito. You can use a variety of sensors or conduct inspections of different types of infrastructure with Scopito. The Scopito application pairs especially well with a highly autonomous drone featuring autonomous mission planning features.
Here are some additional Scopito benefits to your drone inspection operations.
Scopito specializes in visual data management, and we can do a lot to help your project get off the ground. Contact us for a talk about getting started with your data handling the right way.
Enter your details below to continue
Please send me occasional e-mails about the product.
14 Days free trial starts automatically when you sign up