UAS applications promise significant rewards for ambitious innovators, but the question remains, how does the industry accelerate public acceptance of the use of drones to bring about commercial viability? And will the value this brings transform the perception of autonomous robotics more broadly, from that of fear of the unknown to one of trust?

It was asking ourselves these questions that drove us to conceive a whole new concept for autonomous, collaborative uncrewed aerial vehicle (UAV) operations. Our thinking from the get-go was clear: single ship UAV operations just doesn’t cut it.

Benefit for drone operators will only be achieved by progressing towards a team approach that increases synergy, performance, and efficiency of operations. Essentially, we want to improve the way UAVs are used, resulting in decentralization, self-organisation, resilience, operational flexibility and environmental sensing.

We struck on the idea of autonomous collaborative drone fleets with a dynamic hierarchy of command and control, proven as a very effective method across all types of highly complex operations. This is a concept we expand upon in our latest Innovation Briefing – ‘Collaborative autonomous drone fleets for next-level UAS operations’. We believe this model of UAS technology, which we have now developed into working full proof on concept, has the capacity to form an operational fleet of UAVs that is far greater than the sum of its individual parts.

The use of UAVs for commercial applications is rapidly becoming one of the world’s fastest growing industries. Success is predicated upon both commercial viability and public acceptance, which can be accelerated if the ecosystem is seen as reliable, safe and trusted.

UAS drone usage initially began as a three-dimensional extension of human applications, and ultimately added to the cognitive load of the operator. The dream is to advance the industry to be fully automated, integrated and autonomous. This could see autonomous vehicles across all domains, including air, land, sea – and we envision this as a phased journey, with a need to mitigate inherent risk across a range of complex challenges.

The standard UAS mission model is a semi-autonomous drone with autopilot, which can take off, land, navigate and avoid obstacles. This model requires a digital connection and oversight from at least one human operator.

Mitigation strategy typically includes automated action to land or return-to-base if something goes wrong, such as loss of GNSS signal. The current model of deploying homogenous drones, which work identical to and independently of each other is inefficient, lacks adaptability, and is limited in use cases.

Collaborative UAVs to enhance synergy

And that fact brings us to the crucial point – could our collaborative model of UAVs enhance mission execution synergy by accomplishing more in less time, and in a safer manner than the way humans and drones currently do? For example, if a critical component malfunctions, such as a communications link with the ground control station, the fleet identifies another member of the flight to directly link to the degraded drone and provide the necessary information to continue.

Straightaway, the advantages of the model become clear. Complementary and redundant capabilities, an understanding of each team member’s status, and improved robustness means better mission success rates for service providers. Additional safety measures make the model more appealing to regulators and will lead to greater social acceptance.

The concept is highly efficient because the members have discreet roles, responsibilities and varying capabilities. Even the smallest team of two drones in the fleet can realize efficiencies when both members are properly matched and task organised.

In addition, onboard AI can enable each drone to monitor its situation and communicate progress, local insights and health levels to each other. This distributed intelligence enables the drones to respond to others in the fleet, identifying and compensating for risks when any member experiences technical issues or loses communication.

Collaborative fleets have the ability to respond to dynamic environments and changing mission requirements. They can adaptively create scenes of varying fields of view and use AI to create situational awareness from synthetically generated imagery.

There are many use cases that lend themselves to this. For example, drones can spread out geographically to capture images of a scene for infrastructure asset inspection, then create a synthetic image of varying sizes and resolution depending on requirements for the use case at hand. Railway inspection is a good example of where this would offer huge value to an infrastructure owner, as shown in the video below. This distributed intelligence offsets the dependence on complex centralised data processing, reducing latency and energy consumption, both of which are critical determinants of performance.

Beyond visual line of sight (BVLOS)

As one person cannot see the whole picture, a properly organised fleet of intelligent drones vastly enhances situational awareness, obstacle avoidance, beyond visual line of sight (BVLOS) communications range and safety of operations.

Our premise is that this approach is already tried and tested in the sphere of human endeavour. It is proven by decades of in-flight experience, in highly complex environments including experimental flight testing and helicopter-borne shipboard replenishment.

These encompasses a wide spectrum of human physiology, situational awareness and highly coordinated multi-aircraft flight activities. We firmly believe this can be applied to UAV operations, with continued advancements in AI, smart platforms, and extended battery life key to making it all happen.

Our inhouse digital service innovation (DSI) team has been central to the concept development, which we recently showcased at Amsterdam Drone Week. DSI is responsible for the macro integration of the UAV mission into the overall business and human benefit – why a flight takes place, essentially. But DSI also provides the logic of how the fleet makes decisions, while considering the pivotal relationship between data and computer processing, vehicle and cloud, to ensure that the mission succeeds.

Comms is of course vital to allow the collaboration… from fleet to mission control and from fleet to the recipient of the service. We envision a multi-layered approach to ensure resilience using short range, terrestrial and satellite communication interlock.

AI is instrumental in detect and avoid, allowing the drones to sense themselves and negotiate their operating environments. Our teams have developed the algorithms that will cope with the ever-changing complexity and unpredictability of mission environments.

The multidisciplinary team here at Cambridge Consultants is excited by the potential of the concept. The UAV squadron we’re proposing will be a truly multitasking force, doing the work to an order of magnitude greater than humans can currently achieve.

Shared technology loads across multiple drones, one vehicle acting as the relay for another, in-flight charging, multiple drones each with complementary functions, tools as part of a complex task… the list goes on. Read the CC Innovation Briefing explaining CC’s breakthrough work, and please reach out to me if there are any aspects of this topic you’d like to discuss further.

専門家