UAS Autonomy and AI

Author: Dr. Michael Zimmer

_______________________


Autonomy and Artificial Intelligence (AI) in Unmanned Aircraft Systems (UAS) improves efficiency by reducing the need for an operator. Furthermore, autonomy and AI in UAS provides routine, periodic, and real-time information sharing. As autonomy and AI technology evolves collection of data and serves will become be utilized as a resourceful tool by organizations and agencies. Autonomation and AI technology in UAS are capable of variation of applications, thus UAS demand is increasing. Within this Blog, the I will discuss on three foundational principles of UAS autonomy and AI; UAS Levels of Automation, Automation vs. AI, and the Automatic Dependent Surveillance Technology project. Lastly, I will share multiple positive and negative effects when too much of autonomy and AI is applied.

UAS Levels of Automation

Integration of UAS into the NAS need to hold regulation on the different automation levels. UAS usage will come in the form of all forms of automation. A standard/blanket policy to capture UAS automation will be ineffective because of the degree which UAS will/will not depend on an operator. Because of this factor, NAS implementation will continue to be delayed. Below, summarize automation levels that will require separate NAS measures.

Level 0 - No Automation. The operator performs all operating tasks like lift, pitch, airspeed, and so forth.

Level 1 - Operator Assistance. The system can assist with some functions, such as sensor detection.

Level 2 - Partial Automation. A “return-to-home” feature with the operator ready to take control of the system to ensure control and safety.

Level 3 - Conditional Automation. Establishing an autopilot flightpath, for which the system will hit assigned markers.

Level 4 - High Automation. System is capable of all basic functions to include monitoring and will notify the operator for assistance when needed conditions

Level 5 - Complete Automation. Autonomous flying requires no human attention.

Automation vs. AI

The terms of automation and AI are often used interchangeably. Automation and AI share similar traits to include, software, physical movement, and self-machine operation that allows for efficiency (Frank, 2019). However, automation and AI hold different complexity levels in the manner of machine execution. Automation is automatic actions with minimum human intervention through pre-programmed hardware or software. Artificial Intelligence, is intelligent machine actions based on the machine’s choice that supersedes human direction. For example, my toy robot will dance on a voice command as in the movie Terminator, the T1-1000 makes self-decisions to dance when prompt. Automation and AI can occur without one another and it is the programmer’s choice when establishing device propose.

Automatic Dependent Surveillance Technology

Due to the increasing amount of hobbyist, military operations, and recent mail delivery growth, the Federal Aviation Administration (FAA) launched a complete revamp of the National Airspace System (NAS) to hopes of integrating a global positioning system (GPS) technology. The thought behind the FAA project stems from the Traffic Collision Avoidance System found on nearly all commercial and military aircraft. In contrast, the concept is to provide safe separation between all aviation vehicles, conflict resolution, and see and avoid. The proposed newer technology is called Automatic Dependent Surveillance Technology (ADS-B), and it's to begin its solution of determining aircraft position while aiding Air Traffic Controller situational awareness on January 1, 2020. This upgraded system change will bring the NAS into the 21st Century and will help as a supporting structure capability for any future traffic projections. However, ADS-B is costly for many aviation companies and individuals to ensure compliance because the core of the development and implementation stemmed from latent and over-budgeted NextGen systems. Although the conceptual program design is progressing with minor adjustments, skeptics are addressing potential vulnerabilities such as data mining or GPS jamming as a hindrance to the integrity of the technology. Furthermore, most NAS collision avoidance system designs are based on a Distributed Model Predictive Controller for trajectory tracking, meaning, Right of Way rules prescribed by the International Civil Aviation Organization are for human piloted flights (D’Amato, Mattei & Notaro, 2020).

Conclusion

Automation and AI has proven itself to be revolutionary technological advancement which reliably enhances flight operational safety and efficiency. Although automation and AI holds safety and efficiency benefits, pros/cons have to be weighted based on mission criteria and autonomy usage. Assets will be determined on autonomy and AI level capability. Thus, below are a few pros/cons:

Pro 1 - Greater Situational Awareness. Reduce direct operational involvement, crews are able to allocate their concentration towards maintaining awareness with their environment as well as interpersonal communication with their team.

Pro 2 - Operating Cost Reduction. With computer management, the heading, altitude and airspeed whereby the system can travel at the most ideal efficiency

Con 1 - Over-dependence on Automation. Increased reliance on automated technologies, crews may rely excessively on autocontrols and can lead to negligence

Con 2 - Adverse Impact on Airmanship. Crew skill development may be adversely affected by their excessive dependence on the electronic instruments.

_________

References

D’Amato, E., Mattei, M., & Notaro, I. (2020). Distributed reactive model predictive control for collision avoidance of unmanned aerial vehicles in civil airspace. Journal of Intelligent & Robotic Systems, 97(1), 185-203. doi:10.1007/s10846-019-01047-5

Frank, M. R., Autor, D., Bessen, J. E., Brynjolfsson, E., Cebrian, M., Deming, D. J., Feldman, M., Groh, M., Lobo, J., Moro, E., Wang, D., Youn, H., & Rahwan, I. (2019). Toward understanding the impact of artificial intelligence on labor. Proceedings of the National Academy of Sciences of the United States of America, 116(14), 6531–6539. https://doi.org/10.1073/pnas.1900949116

Harner, Isabel. (2020). The 5 Autonomous Driving Levels Explained. Automation. IoT For All. Retrieved from https://www.iotforall.com/5-autonomous-driving-levels-explained/

Popular posts from this blog

Design and Operational Analysis

Image
 Author: Dr. Michael Zimmer ___________________________________________________________________________________ An increasing demand on package delivery unmanned aerial system (UAS) have accelerated commercial UAS innovation. Amazon has invested $130 million in the infrastructure of UAS delivery since 2015 and commented an additional $350 million for operational support (Keeney, 2015). Amazon believes their projected capital will allow Amazon to net a 15% return on $1 deliveries (Keeney, 2015). Package delivery UAS faces mass rollout delay constraints surrounding regulation, UAS configuration, application, and consumer demand as these areas will need to mature. Within this paper, the researcher will examine the UAS application of package delivery while constructing a Octorotor through Embry-Riddle Aeronautical University-Worldwide Hub program. In addition, this paper will discuss on the Octorotor’s package delivery performance within a simulated operation while exploring regulatory...
Read more

Testing Optics and Configuration

Image
 Author: Dr. Michael Zimmer ___________________________________________________________________________________ According to Ding et al. (2020), Unmanned Aerial Systems (UAS) can support specialized optics and payload while vertically taking off and landing. This support of specialized optics and payload have made UAS usage a preferred mean of intelligence, surveillance, and reconnaissance (ISR). Unmanned Aerial Systems optics and payload are generally assembled onto an aircraft based on mission needs. Embry-Riddle Aeronautical University’s Visual Optics/Thermal Lab allows learners to examine UAS optic and thermal performance while adjusting optic and thermal settings. This paper will explore Embry-Riddle Aeronautical University’s Visual Optics/Thermal Lab from a learner’s perspective while examining payload and sensor capabilities on a single rotary wing UAS platform. Lastly, this paper will discuss on payload and sensor configuration onto an UAS, flight planning, and on a simulat...
Read more

Testing C3 and Configuration

Image
 Author: Dr. Michael Zimmer ___________________________________________________________________________________ According to Stansbury et al. (2009), Command, control, and communicate (C3) are essential unmanned aircraft systems (UAS) functions that are established under US National Airspace System (NAS) and Federal Aviation Administration purview. Based on regulatory guidelines and airworthiness standards, manufactures compete to exceed basic C3 capabilities. Embry-Riddle Aeronautical University’s Communication Lab allows learners to explore a similar UAS manufacture C3 Lab. This paper will examine Embry-Riddle Aeronautical University’s Communication Lab from a learner’s perspective while exploring common communication systems that are integrated into UAS design. Last this paper will discuss on communication effects using experimental UAS within operational routes.  Communications Lab A communication lab will typically have a received signal strength (RSS) meter using eithe...
Read more