One of the most common questions we have after people have before completing RPAS Training is: What can I do with my drone and what exactly are they good for? Sure – they are a ton of fun, but to consider them as only toys is to miss the true value of this technology that soars.
In actual fact, there are thousands of practical uses for RPAS (Remotely Piloted Aircraft Systems; the fancy acronym aviation professionals now use for drones). I can’t tell you an exact number because in all likelihood, by the time I’ve finished writing this someone will have created a new one. To better understand the RPAS industry it is important to realise that by itself a drone has little value. They find their niche as a mechanism for extending the capabilities of other equipment. Think: cameras, sensors, sprayers and other payloads. It is this capability that is driving the rapid implementation of drones in the wide range of industries that are finding use for them, and this is where their true value lies.
In this sense, a drone then becomes part of a larger process, something best described as RPAS operations. These operations range in complexity, loosely centred on the sensors attached, and the derived data requirements required at the end (i.e. photograph, video, DSM/DTM). Every RPAS operation is different and as a result has very different requirements. In the real world there is a messy spectrum over which these operations fall, from very basic to exceedingly complex. For the sake of brevity we will divide this spectrum in half, grouping operations as either basic or advanced. Note however, that this is very much arbitrary and your applications in the real world may look very different as it will take into consideration things such as end user requirements or specialised data processing.
So you still may be asking your self what can I do with my drone? Here are a few basic RPAS operations that have been the bread and butter of this emerging industry for some time. These missions can generally be performed with relatively low-end equipment and a single competent operator, though more detailed work can require multiple personnel. Some examples of these operations include:
Drones enable access to locations and perspectives previously impossible to capture with a handheld camera. So you want a selfie of you and your mates on your next fishing trip? Forget your overused selfie stick, try taking one from 15 meters in the air, favourite soft drink in hand, while pulling your choice of
ridiculous charming facial expression. Facebook will love it, and just quietly, it gives you a chance to show off your new boat too!
Albeit entertaining, this is only the tip of the iceberg when it comes to drone footage. The Real-Estate sector is increasingly using aerial photography to give unique perspective on their sale properties. Natural resource managers and farmers can analyse aerial imagery to identify problems in the landscape and there is no better way to show friends your latest outdoor adventure than airborne videography.
In the industrial space, RPAS are particularly useful for investigating the condition of structures and assets that are difficult to access otherwise. Consider the time and resource costs of inspecting the condition of a chimney or stack using traditional methods. This would involve several man hours for a team to barricade off an area, bring in a scissor lift, boom, crane or climb up the structure with harness. In additional to the monetary costs of the equipment, transport and time, there are very real safety concerns that make the job problematic. Alternatively, a competent dual-personnel crew could perform the task in under an hour, without even needing a working at heights permit.
Another growing RPAS application is in their ability to capture additional perspective. This falls under the broad category of aerial survey. Aerial surveys allow end-users to see a location in from a satellites view, something that is surprisingly useful in a range of settings. Property developer’s use RPAS imagery to evaluate and communicate the progress of projects. Farmers and agronomists use aerial datasets to investigate crop health and performance. Similarly, survey is used to monitor the prevalence of invasive species and assess natural vegetation. These are only a small sample of an exhaustive potential list of survey applications. To check out a few of the projects that we have completed over the past few years, have a look at our aerial survey projects and portfolio.
Still not seeing what you can do with your drone, at the other end of the complexity spectrum lies the advanced uses for RPAS. In general, advanced applications utilise larger, more complex aircraft that are equipped with cutting edge sensors, payloads and technology.
A rapidly growing area of RPAS operation involves the capture of spectral data. A standard off the self camera is designed in such a way that it only captures the portion of light that is visible to the human eye. Importantly, the sensors inside the camera capture very broad segments of this light and are thus considered to have low spectral resolution (see: Resolution Requirements). Importantly, modern high resolution digital cameras only have high spatial resolution, not spectral resolution and are therefore not suitable to spectral survey.
Enter the multispectral sensor. This technology is designed to have high spectral resolution in order to generate data sets targeted to very specific parts of the light spectrum. Multispectral systems have dedicated sensors, each covering a narrow band and are thus designed only for specific purposes.
Agriculturalists are at the forefront of these developments, with multispectral sensors finding widespread application as a means of generating NDVI or EVI data sets. This data is then used to identify problems in the crop, well before they manifest visual decline. In doing so, strategic crop management programs can be developed for feeding, irrigation and spraying that address issues much earlier than otherwise possible.
Other industries have also identified the value of multispectral survey. Aspects of environmental management are particularly amenable, including; forest health, soil management, coastal processes, vegetation ecology, sediment flux and erosion control. Geologists have been using multispectral data from satellites for several decades as a tool for mineral exploration. Now that sensors and RPAS have reached a suitable price point, this is readily extending into mine pits for ore grading.
Any given material, i.e. plant tissue, rock, steel, will have a unique spectral signature (think: pattern). Multispectral systems capitalizes on this by sensing only the parts of that signature that will distinguish that material from it’s surrounds (or differentiate status in distinct samples of the same type). For example, NDVI is able to assess crop health as the components of light that are analysed indicate the efficiency of plant metabolic processes. This level of specialized design, while very useful, is the reason sensors are not interchangeable for different applications.
Beyond multispectral capability lies the exciting realm of hyperspectral imaging. Hyperspectral sensors are comprised of a vast array of dedicated imaging components. Historically, this level of sophistication has placed hyperspectral data at a price point beyond all but the most well-financed research institutions. Today, sensors, although still expensive, have reached a point that they are quickly becoming interesting to a much broader market.
Hyperspectral imaging is best conceived by analogy. If standard RGB (visual) photography encompassed a poorly-scanned chapter, multispectral would be a collection of clearly legible highlights taken from several select chapters. Hyperspectral would then be a highly articulate summary covering all the key points within the book. While a modern multispectral might collect information on up to 20 bands, a hyperspectral collects upwards of one thousand discrete bands. These range from visible light, deep into the infrared spectrum.
The applications for hyperspectral data are immense. By providing the ability to investigate a large number of features at once, data sets can be ‘overlaid’ enabling deep insights into process and causation to be investigated. However, the largess of the data also presents a significant bottleneck for the technology, as the processing requirements are hugely demanding.
Advances in artificial intelligence are allowing researchers to develop novel and innovative approaches to hyperspectral data analysis, making this field one to watch. In addition, research conducted using hyperspectral sensors can, and does inform the development of multispectral systems.
Continuing with our book anaology, thermographic imaging returns to the other end of the spectrum. Similar to visible light, thermal imaging involves capturing an entire spectral range. As such, it can be conceived of as a separate chapter in our book. This chapter lies at the far end (longwave) of the infrared spectrum.
Longwave radiation is absorbed by molecules and causes them to vibrate. This vibration is responsible for heat. When you stand in the sun, IR radiation shakes the molecules in your skin cells causing you to feel warm (or stinkin’ bloody hot if you live in North Queensland). This is the reason they are termed thermal sensors – they are in essence a way of measuring a material’s temperature.
Every molecule has a distinct infrared wavelength which it will absorb (and emit) and this is the premise behind thermography. By identifying an object’s infrared radiation emissions, information about the composition of the material can be collected.
Thermal imaging is very useful for a large number of specialized applications. Unlike other spectral methods, thermal sensors are designed to detect the radiation that an object emits, not that it reflects (this is a separate and not necessarily related phenomena). Mechanical systems, certain chemical reactions and metabolic processes in living cells all emit IR. Therefore the behaviours of these systems can often be investigated by thermography.
Equipping an RPAS with a thermal sensor provides this capability at extended ranges and greater physical access. Ecologists use thermal imaging platforms to find and monitor wild and invasive animal populations. Emergency services employ these systems in a range of applications; from delineating fires through smoke, to locating missing persons (or criminals) in vast areas of vegetation.
For mechanical and structural applications, engineers utilize thermography widely as a means of investigating component integrity. Thermal images are also useful for heating and cooling systems, where potential leaks can be readily identified. Similarly, the conductions of electrons in electrical systems generates IR. As a result, faults and failures often display as thermal anomalies.
The list goes on…Another application of UAV derived data is in the development of detailed 3D models. Through the use of aerial photographs and cutting-edge software, detailed models of terrain and structures can be generated. Digital surface models (DSM) and digital terrain models (DTM) are widely used examples. Any industry that requires the quantification or visual inspection of land, structures or materials, including changes thereof, will find value here. Mining, construction, environment, agriculture, tourism, emergency services and government are all active users of 3D models in one way or another.
Of the myriad potential applications for these models, most are under-exploited, many are largely undeveloped and an unknown number are yet to be discovered. Contemporary examples include stockpile estimation in mineral resource recovery, growth analysis in agriculture and rapid land survey in infrastructure development.
An important caveat here is that good data is very much contingent on the GPS capabilities of the platform. Modern GPS in good conditions can obtain relatively accurate position in the horizontal plane. The problem for many applications is the need for strict accuracy in the vertical plane. Without expensive and specialised GPS systems, sufficient vertical accuracy is impossible to achieve.
For applications requiring these high levels of spatial accuracy (centimetre or better), LiDAR is the only viable option. LiDAR systems operate by emitting packets of light and recording the time they take to return, or the change in phase over their adventures. In doing so, they create dense clouds of points, known, ironically, as point clouds. These point clouds are very accurately representative of any and all features in the vicinity of the sensor.
Similar to hyperspectral, LiDAR systems have traditionally been prohibitively expensive as well as power hungry. This is quickly changing with a number of models on the market now available that are suitable for UAVs, and now LiDAR capacity is evolving rapidly. Some examples include forest mapping, internal volumes, terrain modelling and asset inspection to name a few.
So…What can I do with my drone?
That, in short, is what you can do with your drone. There are still many other applications that we haven’t touched on above such as interactive art works, surveillance and cattle round up. In reality, this represents a very limited subset of a much larger list of things that are possible with RPAS. This industry is developing at break-neck speed and there is huge potential for motivated and competent operators to find a niche amongst the chaos. So if you are still wondering what you can do with your drone, you might be about to come with the next new and awesome idea! I would say the sky’s the limit, but that will get me in trouble. According to CASA, 400ft AGL is the limit.
Bad jokes aside, drones are a whole lot more than just toys. If you would like more information about anything in this article, get in touch with us. We are always happy to help.
All the best in your drone adventures!