Learning From Drones: Where Automation Is Headed in 2017

Over the past year, it has become clear that the technological world is mostly on one path - the path to automation. However, anyone who has ever tried to build any kind of automatic device knows that the challenges from creating, to testing, and regulating are all a steep uphill battle. One area that has made some of the most significant strides in automation is the drone ecosystem. Automation was already one of the next steps for drones for a few years now, as we could see with Amazon planning automatic deliveries with drones since late 2014.

Various industries are looking into how to transform their technology and make it automated. When preparing to make this change, these industries can learn from the advancements in drones and automation. Here are the five main challenges facing the world of automation, that the drone community has already faced head first:

1. Regulatory Challenges

For drones specifically, the largest regulatory challenges to automation were presented by the recent rule 107 provided by the FAA in the U.S. This mammoth 624 pager addresses automation both directly, and indirectly through limitations on the Visual-Line-of-Sight (VLOS). To keep it relatively short, many of the proposed rules and concerns regard Visual-Line-of-Sight and Automation. It should be noted that while some exemptions have been granted, to some aspects of these rules, the future of automation absolutely requires the possibility of exemption when appropriate. Proposed solutions have been offered to the concerns expressed. In order to make automatic drone systems safe, I believe that certain steps should be taken, such as: flying in areas cordoned off from people, reliable and tested hardware, rigorous failsafe parachute, return-to-home failsafes, geofencing, and known regular scheduling times of missions. Automated platforms in all areas are going to face similar regulatory challenges.

2. Technological Challenges:

The greatest challenge to building automatic hardware is the integration of many different kinds of engineering professionals and expertise. The development of an automatic system can require the capacity to change batteries and payloads without a human operator - for example in the world of drones. When building anything robotic and automatic, a significant challenge is minimizing points of potential failure in a technologically complex system. This involves finding the right engineering professionals from mechanical, electrical, software and robotic backgrounds.

3. Challenges of Defining 'Automation': "What is an Automated Platform?"

While automation is not overly difficult to define it is talked about quite loosely, and the status of a particular product might be difficult to place. This is, in part, due to two similar sounding words, adopted from Greek, with close enough meanings to invite confusion:

We use the word Automatic (Self-moving, acting on its own), to generally apply to machines who do their job in a preprogrammed way. The term was coined by Heron of Alexandria, in his book Pneumatics, who built a machine to open temple gates after a sacrifice. A system is fully automated when the human component - for drones when the pilot who launches, flies, lands and replaces the batteries of the drone - is able to be removed from the process.

We use the word Autonomous (Self-governed, having its own laws) to generally apply to machines with autonomy: those with a greater or lesser ability to go about the world unguided, perhaps in a goal directed way, and solve problems as they arise. Development of autonomy is much more closely associated with artificial intelligence than automated machines which are linked with production efficiency. What exactly true, or full, autonomy might be is something of a philosophical question. However, a reasonable rule of thumb is the greater or lesser control a human agent has on the machine's behavior.

A challenge to this way of talking is that such definitions aren't mutually exclusive. A single machine might possess elements of both of these kinds of systems. Sense & Avoid (S&A) technologies, used for emergency obstacle avoidance, for example, can be equipped as additional safety measures on an otherwise automatic unit. This means that a fully automatic unit might also be partly autonomous.

4. Challenges of Combining Fields: the Role of Software in Automation?

Software obviously must play an important role for technology to reach full automation. Many times, a drone is thought of as an aerial data collection device, and in this case, the drone harvests the visual data in its most basic form. There is often significant processing, both by humans and software, required to turn this brute data into the actionable insights that a client requires. If automation requires removing the human element from a process, then automated processing of data is essential.

5. Cultural and Marketing Challenges

One of the most interesting challenges, specifically faced by drone automation, is a lack of imagination as to the utility of aerial data. This is a more interesting story than it, at first, might seem. Big industries are, by nature, late adopters of new technology. The bigger the scale of the industry the more a disruptive technology is, well... disruptive. This includes the initial proof of concept of the tool, in this case, drones. Industries typically follow a fairly routine path to drone automation. They start off by buying, or leasing, their own consumer drones - which are relatively cheap - as well as simple payloads/sensors/cameras - which are not. Further expenses include the training and certification of new pilots and outreach to regulatory bodies. With this approach, a lack of expertise is quickly felt, not simply in the potential for crashes, but in limitation of the ways in which the equipment and sensors can be used. For any industry that wants to become an automated platform, it is important to not only understand how the market will change, but understand how becoming automated can open themselves up to new challenges. It is important for industries to invest their time and money into the equipment and software that will make their automation great, so that they don't need to keep putting money into products that won't deliver.

What Are the Next Up-And-Coming Industries in Automation?

While the drone industry is certainly dominating automation at the moment, one up-and-coming industry in automation definitely worth mentioning is the mining industry. The development, implementation and refining of fully automatic mining operations is a great achievement. I hope that drone operations can be a big part of this enterprise. Another equally interesting industry is the maritime industry. Rolls Royce, who have recently revealed a plan to develop automatic vessels monitored from a remote control center, see big potential in automating this industry.

While drones are making significant strides in automation, it is time for other industries to step up to the plate and make their own steps in automation. With this in mind, industries can learn from the drone world how to make it through and surpass different challenges facing them, whether it be regulatory, software, or marketing challenges.