Scaling drone operations with the use of multiple autonomous drones is a complex subject in the industry at the moment, but one that needs to be addressed in order to facilitate the development and expansion of the business cases currently present in the ecosystem today. This includes activities such as industrial asset inspections, surveillance, agriculture, and search and rescue, among others, which can all be improved on significantly with the operation of multiple simultaneous autonomous drones.
During the panel discussion led by Stewart Marsh, who currently leads Cambridge Consultants’ aerospace activities, the topic was discussed by panellists Eszter Kovács, CEO & Co-Founder of DroneTalks; Lori DeMatteis, Chief Growth Officer at MatrixSpace; Amanda Boekholt, Legal Advisor and Project Stakeholder Manager at the Swiss Federal Office of Civil Aviation; and Chris Forster, Chief Operating Officer at Altitude Angel.
The overarching theme of the panel was to explore the diverse potential that the use of multi-drone operations has in modern business cases while illustrating examples of the advantages they offer over the use of a single drone. The benefits of multi-drone operations encompass several dimensions, including increased functionality, improved resilience and reliability as well as faster task completion.
The panel began by discussing the use of drones in the agricultural sector, where more drones mean faster mission completion times for things like field inspections or crop spraying. The discussion also brought another important aspect into play, which is how connectivity is established with the use of drones beyond the visual line of sight (BVLOS). This might mean establishing one drone as the centralised communication hub that will be used to orchestrate a mesh network of the entire cohort to ensure continuous operations even when connectivity capabilities are limited.
The examples cited serve as foundational cornerstones and provide a preliminary indication of the potential benefits associated with the deployment of multi-drone operations.
When deploying multiple drones, the horizon of possible use cases for the industry beings to open up exponentially as collective mission objectives and how drones prioritise the routes towards these begins to open new pathways for successful operations. One panellist described how intelligent algorithms on the cutting edge of technology would enable multiple drones to understand the overall mission objective and operate together on how best to accomplish this under the supervision of a single operator. This form of management has numerous advantages, and it highlights the potential benefits of a collaborative multi-drone environment, unlocking numerous new opportunities and capabilities.
Still, the question of one drone versus multiple drones for operations is a viable option today, and if cooperative multi-drone systems can be implemented beyond standard drone light shows in a practical environment.
In the United States, the recent passing of Code 9113 by the Federal Aviation Administration (FAA) has opened up new opportunities and use cases beyond critical infrastructure inspections. These extend to areas such as fire protection, public safety, flood management and levee inspections. Lori pointed out that the use of multiple drones in these types of scenarios can save lives, protect agriculture and mitigate food disruptions caused by the ever-increasing natural disasters due to climate change. Through the regular use of multi-drone operations, preventative strategies can be created to identify potential problems early on and enable rapid response teams to act.
In addition to these use cases, drones can help to revolutionise firefighting efforts by providing real-time monitoring and surveillance capabilities, overcoming the hours-of-operation limitations placed on helicopters. As a result, the utility of drones extends beyond commercial applications to significantly improve the quality of life for people around the world.
The squadron approach is when a mission emphasises multiple drones that are required compared to a single drone, which has been seen as the next stage of drone operations for the industry. These drone squadrons would have a single operator and organise themselves based on pre-defined mission parameters. The system would theoretically be enriched by the integration of advanced algorithms that provide a wealth of benefits from unlocking new capabilities and expanding the scope of drone operations in their entirety.
The panel discussion also touched on the feasibility of using multiple cooperative drones as opposed to a single drone. While regulations are just starting to catch up with this possibility, some recent authorisation from the FAA has opened a few new avenues towards approval. The cases are limited to purposes that have the potential to save lives, protect agriculture and mitigate food disruptions caused by natural disasters. All of these extend beyond the simple realm of commercial applications but have the potential to have a significant impact on the quality of life in the communities they serve.
One limitation, however, would be the safety element associated with the size of the drone itself. While small drones would have a lower safety risk than larger drones, many operations require the use of relatively large and heavy drones — for example, crop spraying, which requires a large container to hold the liquid during the operation. Small drones, however, could be used for things like pipeline inspections, as Eszter pointed out during the discussion. The challenge lies in scaling up to larger autonomous drones, which becomes more difficult from a regulatory perspective.
Providing the perspective of a Swiss regulator, Amanda Boekholt noted that while there is currently limited demand for multi-drone operations, they could prove beneficial in some areas She also expressed uncertainty about the industry's readiness to embrace multi-drone operations and allow fully autonomous flights. While the components for this advancement are theoretically in place, they need to be put together to ensure a smooth transition to full autonomy. Despite the demand, there are significant hurdles to overcome before this potential can be fully realised.
The suggestion for a multi-faceted approach to the use of multi-purpose drones was raised. An example of this would be in the security and first responder sectors, where coordinated missions alongside rapid responses are fundamental. Another industry that would require highly coordinated drone flights is delivery. Here, drones will need to collaborate and integrate their efforts rather than operate independently on point-to-point missions. Automation is key in this scenario, particularly as it's currently impractical to maintain a one-to-one link between a drone pilot and a drone. Automation could, therefore, enable a single pilot to control multiple drones, particularly in the context of a multi-role fleet. These advances should be pursued simultaneously, rather than sequentially, to accelerate progress.
During this point in the discussion, however, Eszter Kovács, called for a careful evaluation of each use case to determine whether automation is truly beneficial or whether it could exacerbate challenges. While recognising the potential for automating certain elements of drone operations, she recommended a gradual, step-by-step approach rather than an immediate move to automate entire processes. She also warned against the potential risks of rushing into full automation with a large number of drones when current regulations are still struggling to accommodate single drone use cases. Instead, she advocates a measured, strategic implementation over time, ensuring that each step is carefully considered to optimise the benefits of automation.
Some panellists argued that maintaining the status of human pilots controlling one or even multiple drones will be a requirement for the foreseeable future, but some automation will come into the fold to allow the one-to-one ratio of drones to human pilot to slowly increase over time. For example, police might be able to use automation to pilot drones along pre-determined routes to help with surveillance efforts, allowing them to control multiple drones at the same time. First responders, on the other hand, will be controlling the drone as it goes into an unpredictable situation, which will be much harder to automate.
Lori DeMatteis takes a less optimistic view. She is sceptical about the industry's readiness for full autonomy and multi-drone operations. However, she acknowledged the presence of technological advances — either in development or already available — that can facilitate incremental progress. Lori emphasises the need for collaboration between regulators and industry stakeholders to demonstrate the potential of these technologies.
Amanda Boekholt echoed the sentiment that the industry is aware of ongoing developments and progress. However, she highlights a gap in the public's knowledge and confidence in these new technologies. The current focus is and will remain, for at least the present, on maintaining the standard one pilot per aircraft process. Introducing the concept of fully autonomous drones without human involvement could potentially create negative or fearful reactions from the public. Amanda pointed out that while the industry is striving to be ready, the public is largely unprepared for such advances. She advocated for effective communication with our communities as the primary means of bridging this gap.
Eszter Kovács cites a recent report by the European Aviation Safety Agency (EASA), which found that 83% of people are open to new innovations which utilise drones. She suggests that as drones become more commonplace, existing concerns are likely to diminish. The public's main concerns are noise and safety, particularly the fear of drone accidents. Drawing parallels with the automated trains in Zurich, she observes how passengers have become accustomed to autonomous solutions, hardly questioning the absence of a human driver. Eszter argues that there are valuable lessons to be learned from the experience of ground transport. She advocates a productive dialogue with transport officials, road traffic managers, smart mobility experts and others to identify synergies and highlight the benefits of autonomous drones in improving efficiency and reducing road maintenance costs.
On the other hand, Lori DeMatteis claimed that the use of multiple drones in a single mission can actually increase safety and reliability thanks to technological advances. Previously, the ability to use edge computing and integrate intelligence using multi-sensor technology did not exist. This breakthrough is a game-changer, opening up previously unimagined possibilities. Lori explained that her vision would be a scenario where a diverse fleet of drones, each equipped with different sensors and augmented by AI and machine learning, can perform tasks such as identification, classification and seamless communication, all without relying on cloud infrastructure.
Still, how does this vision of the future when looking at it from an unmanned traffic management (UTM) perspective? It’s true that, in some use cases, the probability of mission success may be higher when multiple drones are deployed. Still, it's vital to ensure the safety of the hardware and mitigate the risks involved in allowing a fleet of drones into specified airspaces. As the debate around artificial intelligence and machine learning becomes a larger topic in the field of autonomy, a vision of a highly automated sky is emerging as the answer to current challenges faced by the industry. However, this vision raises questions about the reliance on non-deterministic algorithms to make critical decisions for drones. Historically, safety regulations and governmental approvals have relied heavily on defining logical, predetermined responses that the drone would take in certain scenarios. This approach ensures that the drone's actions follow a logical sequence rather than leaving decision-making entirely in the hands of the machine’s software.
This could present a potential barrier to gaining full public acceptance in the future if the industry needs to begin to rely on this type of autonomy to scale the industry effectively.
The regulatory landscape is continually changing, including the SORA methodology, to include things like the ability to establish new U-space airspaces, the 2019/947 and 2019/947 regulations and others, but even with these regulations, things aren’t as scalable as many in the industry would like.
One of the potential changes that would need to be made, explained Chirs Forester during the panel, is that UTM services need to be integrated with the ability to coordinate and orchestrate airspace activities. Without this coordination and integration, even the use of 10-15 drones simultaneously in a single airspace would significantly restrict their use for others, limiting emergency drone operations like search and rescue or air ambulance operations.
Eszter Kovács continued with this idea and gave the example of EVE Air Mobility, a company that has designed an autonomous air taxi that can carry a pilot and four passengers. Even though the air taxi is capable of operating without the pilot, regulations currently require a pilot to be involved until the foreseeable future. Once these regulations allow for true autonomous flight, however, the pilot’s seat can be removed and replaced with additional passenger seats to help decrease the overall cost of tickets, making them more affordable for end service users.
One of the main problems with allowing for more regulations to empower these operations is the lack of data. Amanda Boekholt emphasised that understanding this data is critical to helping regulators identify what works in the field and how to change regulations to fit scaling operations safely. She hopes for a future where organisations can more openly share their data with regulators so that co-creation towards the goal of market growth becomes a regular best practice.
Chris Forster, however, jumped in to explain that it’s incorrect to think that organisations don’t want to share data with regulators. The real problem there lies with protecting customer data or revealing customer information in a niche market. In these situations, it’s to be expected that organisations prioritise their customers’ privacy over the potential commercial interests involved with sharing data — even if it’s with regulators.
Data has always been integral to the functioning of the industry, but the United States faces a unique data discovery challenge. To illustrate this point, Lori Dematteis pointed towards Iris Automation, which has provided data worth nearly $10 million to the government. However, this information was proprietary, protecting the company's intellectual property rights from exploitation.
Eszter Kovács jumped into the conversation to point out data as a culturally relevant factor for some stakeholders and argued for an industry-wide discussion on the issue to determine the nature and minimum level of data sharing required. She cited the example of Dubai, where data restriction laws have hampered the use of drones by the police due to concerns about data storage and transmission locations. This example underscores the need for the industry to address data management strategies to ensure compliance with local regulations and promote the efficient use of drones.
Different classifications of data are critical, especially in culturally significant cases like Dubai, where not all data can be stored in the cloud due to the sensitive nature of personally identifiable information (PII). The architecture of our systems, be it a drone or a UTM system, should have the flexibility to isolate and localise specific components which can meet the objectives of many customers. However, a blanket rule that prohibits data storage or operation in other geographies could potentially hinder the growth of the industry. It is important to work with the relevant authorities to guide them through this process.
When it comes to the concept of responsibility, there is a need to move away from the prevailing 'piece-part' perspective. This isn’t the case for other industries that hold the operator responsible for car accidents when something goes wrong or a plane crash. Why this isn’t also the case in the drone industry had some members of the panel quite confused because pointing to individual components such as the parachute manufacturer or the software developer didn’t make sense. Lori Dematteis recommended holding the operator responsible, and if other components reveal themselves to be faulty, then this should be dealt with accordingly.
While current operations require a one-to-one relationship between human pilots and drones, it’s still a question from the UTM perspective if the systems can handle a scenario where one person is potentially controlling 10-20 drones on a single mission. UTM systems have been cleverly designed to accommodate the ever-changing drone industry, which is expected to grow exponentially in the future. Because of how adaptable these systems can be designed to be, they could leverage highly scalable features with the reliability of cloud-based architectures which are effective at managing multiple drones simultaneously. To illustrate a scenario, Chris suggested an image of a ground control station where a centralised pilot manages an entire fleet. In such situations, the UTM system treats each drone in the fleet as an individual unit, allowing maximum use of the airspace.
In the case of a light display, for example, where the airspace cannot be shared with others at the same time, the UTM system is able to recognise it as a single unit, while also being able to handle individual drone operations within a larger fleet. From a legal and regulatory perspective, a fleet of drones is essentially treated in a similar way to a single aircraft. This flexibility to manage drones both as individual units and as part of a larger fleet enhances the use of our airspaces.
It was also pointed out by panellists that one of the most important UTM U-space services is strategic conflict detection. This is a feature that ensures no two operations occur simultaneously within the same volume of airspace. As such, a group of drones can fly within a given volume of UTM airspace with the confidence that no other drones are operating within that same volume at the same time. The essential aspect of managing multi-drone operations is to treat each drone within a group as an individual entity within the UTM system. This allows each drone to be controlled independently, either by a coordinating drone, a ground station or individually. This approach is similar to how two independent drones would operate in the same airspace.
When multiple drones are flying together in a single airspace, they need to establish a way to detect potential hazards and perform evasive manoeuvres, which are initiated according to pre-programmed instructions. These evasive manoeuvres can either be group-wide, causing the entire fleet to react simultaneously, or they can be individualised. In the latter case, each drone would respond based on its specific location and the nature of the obstacle. This system could mean that while some drones take evasive action, others may not need to respond at all, depending on their position and the nature of the threat.
In drone operations, it's important to consider how closely the operations of multiple drones are linked. For example, it's very likely that in a light show, the swarm of drones will have an emergency management system integrated into it to facilitate a coordinated response in the event of an emergency. This ensures that the drones descend safely without colliding with each other.
The technology required to manage a single drone versus multiple drones simultaneously is also different. For efficient fleet operations, drones need to be equipped with additional sensors and advanced technology. One potential solution is point-to-point communication between individual drones to enable synchronised operations.
The regulatory and UTM perspectives on the technology don't just focus on one component. The whole system needs to be considered and improved, including elements such as the drone's sensors and other technology, the service it provides, and the overall system structure. A holistic approach is required to introduce full automation. Conscious decisions need to be made today to build systems that will meet future needs, especially as Europe moves towards automation.An example of this forward-thinking approach is a project being trialled in the UK and the Netherlands known as Skyway. A "drone superhighway" concept, Skyway involves the deployment of ground-based sensors along a 65-mile stretch between Reading and the Midlands in the UK. This will provide a ground-based detect and avoid solution as a first step towards automation that does not inhibit manual operations but accommodates both manual and automated drones. Automated drones flying within this corridor will receive navigational guidance from the UTM system, which they will implement based on the manufacturer's defined procedures, whether this involves landing, hovering or diverting.
Want to appear in a DroneTalks interview? See our brochure to learn how!
Sign up for the DroneTalks newsletter to get all the updates!
An Expert's Guide to Drone Terminology: Glossary of Drone Concepts and AcronymsGet your free digital copy