Teaming Autonomous Jets: Hivemind + MQM-178 Firejets
TLDRIn this video, Bobby Holtzner, the autonomy lead, explains the complexities of testing autonomous agents on MQM-178 Firejets. The process involved creating AI pilots, integrating them into Firejets, and coordinating with chase planes in restricted airspace. A key challenge was ensuring safety, as the Firejets had no onboard pilots. The video showcases the aggregation of jets in formation, highlighting successful maneuvers like the lag maneuver. The work demonstrates advanced autonomy, aiming to make these behaviors applicable across multiple platforms in future testing.
Takeaways
- ✈️ The project aims to demonstrate autonomous teaming between two jets in Oklahoma.
- 🤖 Autonomy lead Bobby Holtzner explains the complexity of the technology development for the AI pilot.
- 🔥 The autonomous system was integrated into the Firejets after its creation.
- 🌍 Managing the chase planes in restricted airspace required precise coordination.
- 🛡️ Unlike the X-62, the Firejets operate without a human pilot, adding safety concerns.
- 🎯 Extensive simulations and rehearsals were conducted to ensure safe flight paths.
- 🔄 The Firejets needed to plan their paths based on each other's position to successfully aggregate.
- ⏱️ A key flight feature was the 'lag maneuver' where one jet adjusted its speed to let the other catch up.
- 💻 The autonomy team faced the challenge of random starting positions and lower precision than simulations.
- 🏆 The team aims to develop operationally relevant autonomous behaviors across multiple platforms.
Q & A
What is the primary focus of the project discussed in the transcript?
-The primary focus of the project is to demonstrate autonomous teaming of jets, specifically using autonomous agents (AI pilots) on MQM-178 Firejets without human pilots on board.
What were the key challenges faced during the project?
-The key challenges were: developing the autonomous agent (AI pilot), integrating it into the Firejets, and coordinating complex operations in restricted airspace with multiple chase planes, ensuring safety and precision.
How does the autonomy used in this project differ from previous work on the X-62?
-Unlike the X-62, where a human pilot can intervene if needed, the Firejets are fully autonomous, requiring more thorough safety measures and pre-flight simulations since there's no pilot to take over if something goes wrong.
What are some of the tasks that the autonomous agents needed to perform?
-The autonomous agents needed to take into account their own positions and the positions of other jets to successfully aggregate and plan safe, executable flight paths without human intervention.
How did the team ensure safety during these autonomous flights?
-Safety was ensured through extensive batch testing, rehearsals in simulation environments, and coordination with range control officers (RCOs) to guarantee safe, executable flight paths.
What is a 'lag maneuver' and how was it used during the test?
-A lag maneuver is a geometric technique used to adjust flight positions. During the test, when one Firejet was ahead, it used a lag maneuver to swing out wide, allowing the second jet to catch up, showcasing advanced autonomy.
What made the aggregation process challenging during the test?
-The challenge stemmed from the fact that the jets had random starting positions and had to adjust autonomously to aggregate successfully, without the precision of a simulated environment.
Why was the success of the aggregation significant for the team?
-The success of the aggregation was significant because it demonstrated that the autonomous agents could handle real-world conditions and execute the planned flight paths, a major technical achievement.
What is the long-term goal of this autonomy project?
-The long-term goal is to develop operationally relevant autonomous behaviors that can be deployed across various platforms and continue flight testing on multiple jet platforms.
How does this project push the boundaries of what autonomy is capable of?
-This project pushes the boundaries by demonstrating fully autonomous jet teaming in real airspace, handling complex tasks like aggregation and lag maneuvers, with potential applications across multiple aircraft platforms.
Outlines
🚀 Introduction to Autonomy Project
Bobby Holtzner, the autonomy lead, introduces the project where two autonomous jets are being tested in Oklahoma. The focus of the project involves several key areas: the development of the AI pilot, the integration of this AI with the Firejets, and the complex coordination required for operations in restricted airspace. The challenge lies in safely managing these processes, ensuring the AI executes safe and executable paths.
🛫 Differences Between Platforms
A comparison is made between the X-62 platform, which has a human pilot, and the fully autonomous Firejets. While the X-62 offers a human safeguard, the Firejets must operate fully autonomously. This introduces challenges such as handling initial conditions and the jets planning paths based on their respective positions. A lot of rehearsal and testing was done to ensure the autonomous jets would be able to execute complex maneuvers successfully.
✈️ Challenges of Aggregation
The autonomous jets face a significant challenge in aggregation, where one AI-controlled jet has to join and fly in formation with the other. This process requires complex intercept options and has been tested for months to fix as many edge cases as possible. This paragraph highlights the complexity of these autonomous interactions, with a focus on how the AI manages these dynamic situations.
🎯 Successful Testing and Teamwork
The aggregation during the test was considered highly successful, with the jets performing exactly as expected. The predictability of the AI's movements impressed the team, reflecting months of hard work and testing since October. Special recognition is given to the precision achieved in real-world conditions, which are not as controlled as simulations.
🔄 Impressive Lag Maneuver
A specific moment during the test is highlighted where one of the jets performed a lag maneuver, allowing it to adjust its position by swinging out wide to align with the other jet. This maneuver showcased the sophistication of the AI's decision-making in real-time, impressing both the engineers and those overseeing the test.
🏆 Congratulations to the Team
The engineer expresses deep admiration for the technical achievement of the team, acknowledging the complexity of the task. The successful execution of these maneuvers in real airspace, despite more unpredictable conditions than in simulations, was deemed a major accomplishment, demonstrating the advanced capability of the autonomy system.
🤝 The Future of Autonomous Wingmen
Looking ahead, the team envisions broader operational applications of this autonomy system across multiple platforms. They are working towards deploying these behaviors beyond the Firejets, with plans to continue flight testing on various platforms. The ultimate goal is to push the boundaries of autonomous flight and showcase its capabilities to customers and stakeholders.
Mindmap
Keywords
💡Autonomous Agent
💡Firejets
💡Teaming Autonomy
💡Choreographed Set of Requirements
💡Aggregation
💡Batch Testing
💡Restricted Airspace
💡Lag Maneuver
💡Chase Planes
💡Shield AI
Highlights
Introduction to the autonomous jets project led by Bobby Holtzner, focusing on teaming autonomy for Firejets.
The complexity of the flights involves multiple disciplines, including AI development for autonomous agents.
Challenges include integrating the AI pilot into Firejets and ensuring safe, executable paths without human intervention.
Unlike the X-62 platform, where a human pilot can take over, Firejets are fully autonomous with higher safety considerations.
The Firejets need to consider their positions relative to each other and plan successful aggregations autonomously.
Months of testing were dedicated to addressing edge cases for successful aggregation and intercept options.
The aggregation during the flight test was near-perfect, with both Firejets executing predictable movements.
The lag maneuver performed by one Firejet to allow the other to catch up is highlighted as an impressive autonomous feature.
The autonomous aggregation was a significant achievement, especially considering the less precise real airspace conditions.
The success of the project is credited to the hard work of the programmers and engineers involved.
Shield AI's work could potentially serve as an autonomous wingman in future missions.
The team's focus is on developing operationally relevant autonomous behaviors applicable across various platforms.
The success of this project opens the door to future flight tests on other platforms like those of Kratos.
Interfaces have been designed to make these autonomous behaviors deployable on different platforms.
The project is pushing the boundaries of what autonomy is capable of and demonstrating its potential to customers.
