Self-Flying Drones: Who Will Be in the Pilot's Seat?

Levels_Autonomy_Drones.jpg

As unmanned aerial technology advances, drones will be increasingly able to fly themselves, Autonomously. As the shift to Autonomy progresses, the remote-control “piloted” paradigm will become outmoded, ceding instead to autonomous flight with human oversight.  Just as with the self-driving car, Autonomy is the future for drones.

And yet, despite the major technical advances available today, the industry still has a long way to go before fully autonomous aerial vehicles are ready for urban adoption.

At Aeronyde, we are working diligently to develop autonomous systems for unmanned flying vehicles. We believe that Autonomy is the key to making drones safe and secure, and is critical to the scalable growth of the industry.

To help you understand the different levels of Autonomy, here is a framework describing the different capabilities at each level:

6 - Augmented Autonomy

The UAS has high cognition in addition to autonomous mobility, serving as a functional equivalent of a team-member alongside humans. Predictive analytics feed machine-learning enabled “intuition”. UAS is monitored by remote supervisors, while directly interacting with and consuming mission/project managers’ real-time voice-first instructions. Remote fleet supervisors intervene in cases where machine-based decisions cannot be applied and where human-emotion/ethics is a decision-making factor.

 5 - Adaptive Autonomy

The UASs accomplish entire flight missions with minimal remote supervisor intervention, including modifying flight paths or profiles at any time based on environmental, objective, or other cohort conditions. Fleet supervisors rely on real-time Machine Learning algorithms to make majority of mission-critical decisions.

4 - High Automation

Multi-Agent. The UAS is capable of generating and selecting alternative missions/jobs to accomplish objective(s). The UAS accomplishes objectives autonomously in coordination with other agents’ input, unless human supervisor intervenes; UAS may modify with human approval missions during autonomous flight, based on environmental, objective, or other cohort conditions.

3 - Conditional Autonomy

Single-agent. The UAS is capable of identifying and generating alternative flight paths based on airspace deconfliction rules, weather and system health status. Decisions on alternative mission execution parameters and objective are not made by the system. The UAS makes Pause-Resume-Abort mission decisions and accomplishes selected mission objectives automatically after human approval/instruction.

2 - Partial Automation

Single-agent. The UAV executes predetermined/pre-programmed flight mission/job without making decisions for alternative flight paths or profiles based on environmental conditions and airspace deconfliction rules. The system includes simple failsafes for hardware malfunction, energy level depletion and manual input from pilot (e.g. automated flight between waypoints; Abort mission and Return to launch commands)

1- Pilot Assistance

Single-agent. The UAV executes flight maneuvers with pilots manual input, but has stabilization features along x,y,z axes to augment flight performance, as well as cruise function. Heading and latitude are determined by pilot in command. 

0 - No Automation

Single-agent The UAV has no autonomy.

 

We believe that Autonomous aerial systems will soon make possible a wide array of new use-cases, and provide great benefit to the global economy and quality of life.  We at Aeronyde are humbled and honored to bring forth the responsible development of Autonomous UAS.

The Sky’s No Limit!

 

Aeronyde Team