Opportunity overview for startups & SMBs
Drone technology today is one of the very few technologies that touches upon such a broad spectrum of industry sectors and consumer spaces. This phenomenon will be further augmented by the fast pace of innovation this industry is ushering in. Where it stands today, drone technology seems set to disrupt business models, spur new waves of innovation, and significantly improve business processes.
Historically, drones have been in existence for well over a century now, however the modern drones as we know them today started entering public awareness about 15 years ago. At the time, US military deployed the Predator & Reaper drones to defend and strike against targets in their ongoing fight in Afghanistan. For long, drones were associated with military spying and attack operations.
However, in the last 5 years, especially since drones were made available to public, use of drones in consumer space has picked up. Additionally, governments started granting drone operating permits for commercial use. In 2013 Amazon famously talked about incorporating drones for package delivery. These developments led to businesses exploring drones to improve their operations.
As per the latest estimates by Goldman Sachs, commercial drone market will be worth $13bn by 2020. This, will be the fastest growing drone market segment as businesses are now beginning to explore the possibilities that this technology offers. As per Gartner, the next 2–5 years will be an interesting period for drone technology, as it is expected to expand in use cases and become mature, before eventually beginning to plateau.
The time seems ripe for businesses to take advantage of drone technology in multiple ways. As per Business Insider, some of the industries that are currently using drones in a big way include: real estate, construction, mining, agriculture. With time and as technology continues to evolve rapidly, the innovations will further enable more industries to use drones.
At the other end of the spectrum are businesses servicing the drone industry. These, can be broadly divided into 3 sub-segments. First, are drone hardware manufacturers. Second, are drone software producers who create software solutions that are used among other things for flight management and processing and analyzing image & geospatial data that drones produce. Finally, as TechCrunch highlights there is an emerging space of Drone-as-a-service (DaaS).
For startups and SMBs this presents a unique opportunity to review their business models and see how drones can help improve their businesses. Further, these developments offer great options to look at entering drone industry by creating specialized solutions or providing services for large corporations and governments.
What’s driving adoption
There are several things that are driving wide spread adoption of drones. First, is an ongoing drop in prices of drones especially at the lower end of product spectrum. This phenomenon is very similar to the increase in sales of Personal Computers (PC) in its early days. And, it was further supported by fact that the more the PC units sold the lower the prices went, and that seems to be the case with drone today.
The second and probably the most important factor contributing to increase in drone sales is the fairly short duration of new features introduction that leads to opening new areas to drone use. The drone technology itself has been at the forefront of some of the latest innovations, but it has also benefitted from quickly infusing the latest technological advancements from other areas such as in sensor and imaging technology. These independent technology spaces are having lots of activities and their output will further contribute to the growing significance of drones.
Drones are a broad category of products. At the most basic level a drone is an Unmanned Aerial Vehicle (UAV). However, this definition can be further broken down into Piloted and Unpiloted UAV. A Piloted UAV has a Pilot and a Base Control Station on ground from which the UAV is controlled. An Unpiloted UAV is basically an autonomous UAV with the on-board flight controller piloting the drone.
The definition of drone is further set to change as new uses and ways in which drones are deployed are changing. Experiments are on to test drones for underwater and ground navigation applications rendering it usable for applications other than aerial. Use of drones as aerial medical emergency vans is being assessed. If approved, that will further obliterate the traditional meaning of drones from being unmanned to manned.
In terms of product construction, drones can be divided into three categories. First is the common category of multirotor drones. Within multirotor category, the most common drone is the quadcopter (4 rotors), but the hexacopter (6 rotors) and octocopter (8 rotors) are also regularly used.
The second category is the fixed-wing drone. This drone looks very similar to an aircraft. The final drone type is the hybrid which is a combination of multirotor and fixed-wing.
In terms of usage, multirotor drones are mostly used in consumer space. These are generally cheaper than the fixed-wing drone. On the other hand, the fixed-wing drones are usually expensive are sturdier and used by militaries. The hybrid combines the attributes of both drone types.
In terms of navigation, the fixed-wing usually requires a runway to take off, whereas the multirotor has a vertical lift and doesn’t require a runway. The lift of course affects usability. When a horizontal straight line steady drone movement application is required perhaps a fixed-wing is better than multirotor. However, when its required to hover over a place for a while then multirotor comes in handy.
In terms of flight time, some military-grade fixed-wing drones can last for few days, whereas most multirotor drones have a flight duration of approximately 30minutes. Along the same line, a large military drone can easily cost up to $50m whereas a cheap multirotor drone can come under $100.
Critical components of a drone
Flight stability of drone is achieved by a complex set of electronic components on board. At the heart is the Inertial Measurement Unit (IMU). The IMU measures even the slightest change in acceleration and orientation of drone and auto corrects drone’s movement to ensure that it continues along a pre-programmed flight path.
These measurements happen over 100 times within a second thus ensuring that the drone flight is stable without jerks. Inside the IMU are 3 key components — Accelerometer, Gyroscope, and Global Navigation Satellite System (GNSS). Sometimes an additional Magnetometer is also included in the IMU.
The Accelerometer as the same as a tilt sensor we have in several modern consumer electronic products such as smartphones. The accelerometer indicates change in acceleration. A gyroscope on the other hand indicates drone’s orientation change along any of the 3 axes. The GNSS is usually a combination of GPS (US Satellite) and GLONASS (Russian Satellite) systems, and indicates the coordinates of a drone’s geolocation at a given time. Combining these 3 inputs, the IMU steers the drone smoothly.
What makes this navigation system a wonder is the Micro Electro Mechanical System (MEMS). MEMS are small electronic components and are made of conventional electronic fabrication material, the silicon. The MEMS are designed in a way that even the smallest movement makes the fine silicon mesh move and the resulting capacitance change is measured and used to readjust the orientation and acceleration using GNSS coordinates.
Most IMUs these days have a six-axis gyro which is nothing but a combination of a 3D accelerometer and a 3D gyro in a compact unit. These two together help provide more accurate measurements and faster response time.
First person view (FPV) camera
As the name suggests, a FPV camera is an inflight camera that continuously relays live feed of the view around, from the drone as if a person were on board and viewing it themselves. This is an extremely useful technology especially since the pilot is on ground and trying to maneuver the drone.
This field has seen some major interest from drone enthusiast. One key reason for the interest has been the integration of Virtual Reality (VR) based Head Mounted Display (HMD) devices into the drone ecosystem. These devices create an immersive experience for the pilot on ground controlling the drone.
While the FPV technology is really at the cutting edge, it is still not perfect. Two key challenges remain. First, there is latency between the time camera captures data and when it is received on ground. Lowering this latency will really help. Second, increasing maximum range from which data transfer can effectively happen will be crucial. Some newer drones have incorporated 4G/LTE modems onboard to overcome part of the issue.
The range of sensors that drones use is really wide and more are getting incorporated especially as sensor technology is becoming cheaper and more widely available. Broadly speaking sensors serve two key purposes in a drone. First, to help generate rich 3D images, landscapes, and Digital Elevation Maps (DEM).
In addition to generating rich imagery, volumetric land data, these sensors are used in agricultural for examining the health of crops. Some of the commonly used sensors include: Light Detection and Ranging (Lidar), Multispectral, Thermal, and photogrammetry. All these sensors work on more or less the same philosophy for image generation, and are used in conjunction to produce complete images.
Lidar uses captured reflected Laser waves to generate images, something similar to how Sonar using reflected sound waves or Radar using reflected Radio waves. Multispectral imaging uses several light spectra (sometimes over 100 spectra, as in Hyperspectral imaging) to generate images, and Thermal or Infrared sensors use Infrared light to generate images.
The second purpose for which these sensors are being adopted, is for quickly and accurately mapping the surroundings of a drone in real-time while the drone is in flight. This is very critical as the navigation of drone especially in autonomous mode largely depends on being able to accurately assess its surroundings.
Robots extensively use the Simultaneous localization and mapping (SLAM) technology to map its location in real time and accordingly adjust its movement. Drones use SLAM along with Lidar, Ultrasonic, Infrared, parts of computer vision, and Time of Flight (ToF) sensors to accurately map its location.
The ToF sensors measure the time elapsed between the signal emitted and when it is received back by the sensor. This is used in recreating the depth of objects and surfaces around the drone. In addition to this, drones use Altimeter Technology which provides good measurement of altitude and helps adjust and maintain same height while flying over uneven topography.
The use of sensors has not only turned drones into sophisticated flying objects, but it has also opened up new uses of drones and highlighted areas that can benefit from using drones. With more sensors getting added to drones this trend will only grow.
Another technology development that has helped drone industry gain popularity is the videography technology. Video making technology has been undergoing product and technology enhancements on a frequent basis. Drone manufacturers have been quick to adopt this technology into their products, leading to drones now being increasingly deployed for aerial videography, 360° filmmaking etc.
Increases in camera resolution, digital & optical zoom, shutter speeds, all make it possible today to create sharper videos that were not possible just 2–3 years ago. High Definition (HD) and 4k resolution cameras have started becoming available at much more affordable price points.
Drones, especially the high-end ones now have the ability to carry professional level videography equipment on board. This, is leading to drones being increasingly used for shooting videos that would otherwise require expensive helicopter flights with professional camera crew.
Gimbal is an ancient mechanical equipment that has found use in drone industry. Gimbal is a special instrument that maintains the horizontal level of an equipment that is mounted on it. Example include that of a compass on a moving ship, in which case the compass maintains its horizontal level despite ship’s movement along its pitch and roll directions.
Modern day gimbals use electronics to replicate the gimbal affect and provide stability. Most high-end cameras are mounted on a 3-axis (for Yaw, Pitch, and Roll adjustments) gimbal allowing for perfectly stable shots despite drone’s movement and shake. Recent camera technology developments have led to the creation of chips that now have electronic image stabilization feature, thereby eliminating the need to have gimbals altogether.
The drone industry is at the forefront of innovation. As more features get added more areas open up for drone use, and new spaces needing further innovation become evident. With big businesses such as Amazon pushing for relaxing laws around flying drones, commercial drones seem headed towards autonomous flying in a not so distant future. There are at least 2 areas that are focus on intense ongoing R&D and any success in them will immensely benefit the industry.
Obstacle detection & avoidance and collision prevention
Autonomous flying has some of the similar challenges as that of autonomous driving. So, obstacle detection & avoidance and collision prevention, is an important area of ongoing research. There are few high-end drones that use a combination of sensors such as Solid-state Lidar, Sonar, Infrared, and ToF to tackle obstacle detection and collision prevention. Yet, the solution is far from perfect.
With time and with more computing power, and improved Machine Learning & Computer Vision, the sensor output will further improve. As we have witnessed previously how technology innovations in other fields have benefitted drone technology, it seems evident that autonomous driving innovations will surely make flying drones safer.
One issue that affects drone flying is its dependence on GPS waypoint navigation technology. The technology allows a drone to map its entire flight path start-to-finish through a series of georeferenced waypoints. The drone continuously accesses these per-set waypoints through a GPS and reaches its destination. This is where the challenge lies.
GPS signal is not always good in all locations, especially when the drone is flying among high-rise buildings or indoors. The SLAM technology used in robotics does solve some of the problem but this is an area of immense R&D. The key reason for research focus in this area is the potential use of drones for indoor applications such as tracking warehousing inventory etc.
What will impact further usage
What is stopping a much wider use of drones is the current state of regulation. In certain sense the drone industry today is similar to the early days of car industry in the 20th century. Increased adoption of cars led to regulation of the industry. We are witnessing similar trend with drones. Countries are at different stages of regulating the usage of drone for both personal and commercial purposes.
The current US regulation, which is up for change, mandates that drones be flown by a pilot only within their direct line of sight, vastly restriction the drone movements. Secondly, drones are not allowed to fly over the height of 400ft. Whenever these regulations are relaxed, the drone technology will propel significant huge improvements in capturing data.
With the ability to fly at higher altitudes, drones will be able to capture images of large area using same sensors and in the same time, thus increasing efficiency of each flight. Simultaneously, with sensor technology itself improving, higher and longer drone flights will give much better insights.
Battery life is a major issue today with most drones having an approximate flight time of just 30 minutes, which can be even less depending on payload. Currently drones use Lithium-Polymer (LiPo) battery. With intense ongoing R&D in autonomous vehicles batteries, any breakthrough will surely impinge upon drone battery improvement, increasing the flight time.
The distance over which the ground controller can navigate the drone is limited today. There are experiments ongoing to test 4G/LTE for improving this. Any improvements in this space will surely have a significant impact on drone flight duration.
Today image processing has two issues. First, the orthomosaics technology, which lets hundreds of pictures of the same area taken from different angles, be stitched together through sophisticated software to create a rich and accurate 3D image, is really good. Yet, further improvements will make this even better.
The second issue is the image processing time. Currently, it can take days to stitch images into a 3D image. Reduction in processing times to hours and eventually no time lapse between the camera recording live images to producing 3D images, will significantly improve efficiency.
Drones in essence are flying computers and similar to computers drone are susceptible to dronejacking (hacking of drone). Given the amount of data that drones capture, efforts on maintaining data privacy and securing drones will see major activity in the coming years.
Drones are placed at the intersection of multiple highly innovative technology developments. And, more technologies are getting added to drones making them more powerful almost every month. Businesses should take advantage of this phenomenon as it is unfolding, and plan to participate in it soon and in ways that best suit their business needs.
Depending on the type of business, drones can either be used directly to serve business needs such as those of mining, land surveying, construction, agriculture etc. Or, with newer technologies getting added to drones, newer areas are opening up and businesses can take advantage of this to improve their business operations.
Startups and SMBs can benefit from the increasing usage of drones by supplying solutions to improve the services that drones offer. These include developing specialized software to analyze and process data captured by drones. Today mostly visual data is being processed, but overtime with autonomous drones coming into mainstream, data solutions pertaining to Computer Vision, Machine Learning, Artificial Intelligence will all be part of the mix.
Finally, startups can benefit by offering specialized services by using a fleet of drones through Drone-as-a-service (DaaS) model. The possibilities for businesses to benefit from drones are increasing by the day. Drones are set to revolutionize and change the way businesses have been done traditionally. Keeping a close look at the pulse of the industry and preparing to enter it soon will serve startups & SMBs in the longer run.