In the United States, the commercial promise of unmanned aerial vehicles remains constrained by the Federal Aviation Administration’s prohibition on flights beyond visual line of sight (BVLOS) without special waivers. Current waiver-dependent methods—such as visual observers or chase aircraft—are costly and impractical for scaling operations like drone delivery, which companies including Alphabet’s Wing, Amazon, and UPS have been preparing to deploy. The FAA’s measured “crawl, walk, run” approach to integration has slowed progress, forcing some BVLOS-dependent businesses to close before regulations and technology align.
The FAA’s long-standing “see and avoid” mandate for pilots applies equally to drones, meaning a remote pilot or visual observer must maintain direct sight of the aircraft. To change this, the agency has been testing onboard detect-and-avoid (DAA) systems through its UAS Integrated Pilot Program. One milestone came when the Kansas Department of Transportation conducted a nine-mile BVLOS transmission line inspection using Casia, a 300-gram, sub-$10,000 DAA package from Iris Automation. Casia employs an off-the-shelf camera and processor, combining geometric, deterministic algorithms with machine learning to detect and classify airborne objects. “From when we started building the technology, we were very intentional about not just having one big AI algorithm, one end-to-end deep neural net that controls the whole thing,” said Alexander Harmsen, CEO and co-founder of Iris Automation. “We have a mix of geometric, deterministic algorithms that rely on pixels and tracking — sort of traditional computer vision — together with more machine learning, AI to be able to add semantic understanding [and] things like classification systems.”
Harmsen emphasized that regulators prefer systems whose decision-making can be examined in detail. Casia’s limited “dictionary” of reaction maneuvers enables exhaustive simulation and flight testing, an approach mirrored in Amazon’s Part 135 certification filings. Other developers, such as ModalAI, integrate computer vision as standard but allow expansion to sensors like radar. “ModalAI… has a lot of expandable [inputs and outputs] to integrate other sensors such as radar,” said CEO Chad Sweet.
Beyond onboard sensing, unmanned traffic management (UTM) systems are emerging as a complementary layer. Platforms from companies like AirMap and Altitude Angel aggregate data from ADS-B-equipped aircraft, ground-based radar, drones, and hyperlocal weather feeds to monitor low-altitude airspace. “In dense or controlled airspace, automatic deconfliction provided by [UTM services] … will help airspace managers ensure safe routing of low altitude traffic,” said Ben Marcus, co-founder and chairman of AirMap. He noted that BVLOS operations under UTM supervision are already active in several countries and in North Carolina. Marcus sees DAA and UTM as complementary, with redundancy in some cases and selective use in others.
The FAA’s planned remote identification framework is expected to feed additional tracking data into UTM systems, though not without controversy. DJI has voiced strong opposition to mandatory nationwide tracking. “We strongly disagree with the idea that the future of drones involves every drone automatically being tracked in a nationwide database and stored until the end of time,” said Adam Lisberg, DJI’s head representative for global policy and regulatory issues, citing privacy and security concerns.
Reliable long-range communications remain another BVLOS challenge. Cellular and satellite links can extend operational range, but drones must remain safe even if connections fail. Harmsen noted, “Our system doesn’t need an active communication system to link back to the cloud.”
Some companies have sidestepped U.S. restrictions by operating in less congested airspace abroad. Ontario-based SkyX, for example, conducts long-range pipeline inspections with a tail-sitting fixed-wing eVTOL capable of 65 miles per charge. “Since we only operate on Rights of Way, we never fly the drone over any public areas, as a result we are far less likely to have an incident with a person or other vehicle,” said Jason Braverman, SkyX’s CTO. However, Braverman acknowledged that existing DAA technologies fall short for their high-speed, low-altitude missions, underscoring the need for 360-degree situational awareness.
Industry analysts agree that onboard DAA will be essential for fully autonomous, waiver-free BVLOS. “To implement true, waiver-free BVLOS, an onboard system for avoiding other aircraft will almost surely be a requirement,” said Michael Blades, vice president at Frost & Sullivan. He pointed to the variety of sensing modalities—LIDAR, RADAR, electro-optical/infrared—still under evaluation, with standards likely to emerge only after extensive testing.