![drone traffic management systems](https://dronelife.com/wp-content/uploads/2023/11/flying-a-drone-at-dusk-in-the-city-unsplash-431f0d-300x225.jpg)
Study finds autonomy software needed in future drone traffic management system
By DRONELIFE Features Editor Jim Magill
As drone use scales up in the long run, creating an increasingly crowded airspace at altitudes below 400 feet, a recent study by researchers at Johns Hopkins University suggests that increasing the extent of autonomous operations help might create a safer air traffic management system.
The study, published within the Institute of Electrical and Electronics Engineers finds that “the most effective option for achieving airspace safety on account of the expected levels of congestion is probably going by replacing the human-in-the-loop operations with autonomy.”
Experts predict that by 2035 there shall be 65,000 UAS takeoffs and landings per hour. Currently, the busiest U.S. airports can only handle 300 business aircraft operations per hour, which suggests that a brand new traffic management system have to be devised to accommodate the explosive growth in drone traffic.
The FAA has proposed an idea of operations for drone traffic management, but this idea relies an excellent deal on human control of drones.
“It’s not feasible for these processes to scale to support 65,000 operations per hour. So, we’re going to should depend on autonomous operations,” Lanier Watkins, considered one of the lead authors of the study, said in an interview.
Watkins, a senior cyber research scientist on the Johns Hopkins University Applied Physics Laboratory (APL) and chair of the university’s EP Computer Science and EP Cybersecurity programs, said the research team performed a series of experiments to find out how autonomy algorithms can contribute to safety in congested airspace operations.
Amongst other lines of inquiry, the team investigated how autonomy algorithms react in “noisy” conditions that reflect real-world conditions in a busy airspace and whether the airspace safety promoted by the autonomy algorithms can be negated by the behavior of “rogue” drones operating in that airspace.
The researchers also performed experiments to discover what sorts of airspace risk the usage of the algorithms could impose.
“The role for ensuring autonomy is to make sure that these autonomous algorithms work properly, that they don’t come across failure states and begin making incorrect decisions, after which small air collisions start occurring,” Watkins said.
“It’s like a double-check on the algorithms, like looking over the algorithm’s shoulder, attempting to be sure that that they don’t make the airspace dangerous,” he said.
Of their study, the team examined the feasibility of making a UAS traffic management (UTM) system that relies heavily on the semi-autonomous operations of drones to soundly transit the airspace and avoid mid-air collisions.
“We have a look at this from an end-to-end perspective, where UAS operators wish to interact with the UTM system to have the opportunity to soundly fly their UAS to deliver products to their customers, and the UTM system manages the airspace and monitors the UAS for conformance to the planned deconflicted flight paths the system provides UAS operators,” the study states.
As well as, each drone operating within the system avoids collisions with moving obstacles using its own collision avoidance software.
The study highlighted the cooperative relationship in the present air traffic management system between the drone operator the UAS Service Suppliers (USS), which comprise a select group of firms approved by the FAA to supply Low Altitude Authorization and Notification Capability (LAANC) services.
“Through the UTM flight phase, the distant pilot on top of things and the USS each are sent data from the UAS, akin to distant ID messages and flight telemetry data. This enables the UAS service supplier to perform conformance monitoring by comparing the UAS’s live telemetry data against its planned flight path and confirming it’s inside bounds,” the report states
Of their study, the researchers added 3D evaluation, “noisy” sensors, and collision avoidance algorithm assurance via airspace risk assessment to the prevailing system.
Additionally they performed a Monte Carlo simulation, taking a look at tons of of 1000’s of various scenarios to predict the probability of various outcomes in cases where there’s a possible for multiple random variables.
This simulation provided three layers of separation management — flight planning, scheduling and collision avoidance — together with various safety and efficiency metrics akin to small near midair collisions and real-time risk assessment.
“We found that within the scenarios that were checked out, these algorithms worked marvelously,” Watkins said.
The study found that each strategic deconfliction and conflict avoidance algorithms “contribute to airspace safety by lowering collisions and negating the consequences of rogue UAS.” The team’s work was based partly on earlier studies that found that one side effect of the usage of autonomous systems was delays in mission completion time.
As a part of its research, the team built a “fuzzy inference system” that uses so-called fuzzy set theory to map inputs to outputs. “Given a certain input, only certain outputs are acceptable,” Watkins said.
The study’s authors acknowledge that autonomy is just not “a silver bullet,” and that some autonomy algorithms might produce unknown failure states which will would make them unfit to be used in an air traffic control system.
The university’s APL has been working with the FAA on similar projects for several many years Watkins said. “So, plenty of these findings have already been shared with the FAA in various ways.”
Although the FAA’s concept of operations (ConOps) for a drone traffic management system doesn’t favor any specific implementation, “it does speak to the philosophical architecture obligatory to supply the services for airspace management,” the study states.
“In point of fact, future airspace services shall be implemented by a mixture of government, industry and standards development organizations.”
Read more: