Historically, the unmanned aircraft system (UAS) market has been dominated by military applications from surveillance to active engagement. In the near term, the market’s military focus is expected to continue. However, with projected growth from around $6 billion today to nearly $12 billion in 2023 (according to the Teal Group), large-scale commercial adaptation is not only possible, but likely.
On December 30, 2013, the FAA announced six U.S. test sites for the development of commercial drones and support systems. These sites will lay the groundwork for a new set of rules and operating parameters for UASs in and around commercial airspace.
This announcement and initiatives from companies such as Amazon have caused speculation about the different types of operations that could be enabled by drones—from food and flower delivery to high-speed police surveillance. However, a full array of technological and operational roadblocks currently prevent mass commercial adoption of UASs.
Aircraft technology: Autonomy vs. remote piloting
In today’s intelligent machine world, the line between algorithms and the brain has become blurred. For years, the most sophisticated UASs have employed “autonomous” algorithms that allow them to operate for hours without human intervention or control. This autonomy can be perceived as true intelligence but also as dangerous—raising concerns about accidents and moral and legal accountability.
Although autonomy is alluring to commercial users thanks to lower personnel costs, it is unnecessary and overcomplicated for most near-term commercial applications and may inhibit regulatory approval and public acceptance.
Researchers from Jabil Defense and Aerospace Services and UAS SafeFlight believe that for the next 10 years, “human-in-the-loop” aircraft that use autonomous systems to alert the operator and allow the person to take control will have the best chance of winning over regulators and the public alike.
Although less elegant than fully autonomous systems, this hybrid approach is significantly more palatable and pragmatic as an initial step. This is not to suggest that full autonomy is impossible. However, the researchers consider a UAS launching, delivering a drink or supporting a police officer and then returning to base without human intervention, to be a scenario beyond the 10-year timeframe.
The UAS designs most likely to be certified for flight in urban areas will use this hybrid approach. Therefore, highly reliable communication between the drone and the human operator, using a system able to prioritize traffic, is a critical concern. However, current cellular bandwidth is not adequate and existing protocols for line of sight (LOS) systems are not scalable.
Modular scalability and standardization
In the airframe and propulsion areas, UAS developers can draw insight from the standardization of military and commercial aircraft. Currently, there are more than 120 different unmanned aerial vehicle (UAV) models with varying payloads, propulsion systems, and power sources. By comparison, there are less than a handful of new fighter jets and only a few standard commercial aircraft models, with only three engine manufacturers.
Standardization of UAS hardware is essential for cost-effective manufacturability that will help drive mass commercialization. The future will depend upon modular components that are fit for purpose, aligned with industry standards, and easy to scale up as demand increases. However, the UAS manufacturing landscape is currently fragmented and not standardized.
Today, the largest gap in fit-for-purpose is in UAS propulsion systems. Smaller drones are operating with adapted model airplane engines, either gas or battery powered. These engines were not designed to carry valuable goods and lack sufficient reliability for operating in commercial airspace. In the near term, the authors anticipate the emergence of a liquid fuel engine that is based on existing technology but designed specifically for UASs and able to pass engine reliability tests similar to those for commercial engines.
The myriad of missions expected to be performed by UASs will require modularity to drive down costs. A single UAS platform and engine must be able to complete several different types of missions to provide economic benefit to the end user.
For example, to justify investment in a UAS, a municipality needs to be able to configure the same vehicle for different uses, such as speed monitoring, first responder support, and crowd surveillance, using modular kits. In a commercial example, a business delivering goods may need to modify the hardware to accommodate different payloads such as fragile flowers, perishable food, or bulky books. These modules should be quickly and easily interchangeable.
To achieve modularity and cost-effective, scalable production, the industry must draw on models from the commercial aviation and consumer electronics sectors, which are based on global standards and produced using consolidated global manufacturing, rather than disparate assembly operations at the level of a hobby shop.
Historically, U.S. DOD drones have relied on line-of-sight (LOS) solutions or satellite communication (SATCOM) approaches. Although these point-to-point solutions have proven adequate so far, primarily because the number of UASs operating in any one geographical area continues to be quite limited, they will not meet the needs of commercial drone operations.
Currently, it would be difficult to find any 5 mi² area on earth in which even 10 UASs have operated at the same time. However, by 2018, 7500 small drones are expected to be in the air over the United States.
Networked mobile technology and connectivity are expanding exponentially. Using networked mobile infrastructure as the backbone of UAS piloting communication will be essential to support growth of commercial drones. To this end, networked mobile protocols and approaches must be adapted to meet the connectivity needs of a high concentration of UAVs in flight.
Solutions at scale
To address the above issues, the authors propose efforts across three main areas: adoption of standards for propulsion and power solutions, adoption of open architecture modularity to drive fit for purpose, and adaptation of mobile communications protocols.
• Industry-standard propulsion hardware
The combustion engine has been a reliable solution for nearly 100 years. Despite several billion dollars invested in batteries and hybrid power solutions, very few technologies have come close to the energy density or cost-efficiency of gasoline/diesel engines. Existing small liquid fuel engines (two-stroke technology) can provide a viable solution for commercial applications, providing on-station time surpassing that of batteries and high reliability in operation. The FAA should select two to three engine designs and require a Federal Aviation Regulations FAR 33 or similar test to prove their reliability. With two to three approved propulsion solutions to work with, UAS designers can focus on airframe variations.
• Modular mission kits
Scalability of UAS manufacturing requires basic architecture with interchangeable, mission-specific modules, such as cargo payloads, sensors, and transmitters, rather than separate, customized designs, akin to today’s computers and smartphones. UAS manufacturers should focus on systems architecture as well as software and data algorithms, while leveraging existing electronics suppliers—which have the required expertise and infrastructure—to cost-effectively manufacture the modules and their electronics. It took 500 years for the shipping industry to adopt the modular container, but under 20 years for personal computers to become interchangeable. It is expected that UAS designs will become common and modular in even less time.
• Communication over mobile technologies
Commercial UAS command and control systems will need to leverage multiple personal communications service (PCS) connections through multiple streams. These communications are likely to utilize existing carriers and their installed cellular infrastructure and towers and will require from a few to more than 20 simultaneous cellular connections. Multiple connections will ensure both safety and redundancy across different UAS classes and missions. For example, a mission that requires continuous video feed at high speed will demand more connections, whereas a simple delivery may only need a few.
In a scenario where a UAS mission requires 12 simultaneous connections in an area served by three unique cellular carriers (e.g., Verizon, ATT, and Sprint), hardware and software must balance bandwidth for four connections each across the three providers. For efficacy and redundancy, the system must drive connections with each carrier across a unique tower. In most cases, an UAS with an altitude of 100 m will naturally distribute the four connections among available cellular towers, but must be coordinated through purpose-adapted software and hardware.
In addition, the hardware on the drone needed to establish multiple connections must be effectively managed and size, weight, and power (SWAP) must be considered. Modern manufacturing and the latest electronic advances will be required to effectively make the trade-off between minimizing SWAP and maximizing connections. Once custom hardware utilizing multiple compact cellular data solutions is tested and certified across platforms, up to 20 simultaneous connections for commercial drones will become a reality.
Under the assumption that a human operator must be able to take control of an UAS at any time, command and control technology is of paramount importance. UAS command and control is bi-directional: the UAS must send operational sensor data (position, avoidance system detections, etc.) and the human-in-the-loop must be able to send control data (e.g., turn left, maintain altitude) to the UAS. Latency in the network system must be minimal or it can adversely affect safety and operability.
However, the overall amount of data being sent, in either or both directions, is (compared to the capacity of the system) rather small. Even when the human-in-the-loop is operating the UAS, low data rate video, on the order of a few frames per second, is sufficient. With the exception of low data rate video, command and control data comprises a few thousand bits per second.
The most simplistic approach would be to send all command and control data over all available paths. While this is not bandwidth-efficient, the trade-off (in simplicity of implementation and practically guaranteed delivery of the data) makes it a good option. However, to be viable, current bandwidth software must be developed to account for duplicative receipt from multiple channels with very high efficacy and low latency. This type of software exists, but needs to be adapted for UAS purposes and subsequently tested and certified by the FAA.
Overall, the future use of UASs in myriad commercial operations is highly likely. How rapidly this becomes commonplace depends on the ability of UAS operators, designers, and regulators to reduce barriers and align on key technology goals. Existing communication, propulsion, and manufacturing technologies can and should be rapidly adopted for mass commercialization of UASs. However, adoption will require industry collaboration and agreement, potentially in conjunction with FAA test sites and certification.
This article was written for Aerospace Engineering by Scott Gebicke, President, Jabil Defense and Aerospace Services and Tim Krout, Founder, UAS SafeFlight