A Lunar Cargo Lander with an integrated ISRU Based Oxygen Plant and Autonomous, AI Based Operations
- Paper ID
5589
- author
- company
EADS Astrium Space Transportation GmbH; AI: Aerospace Institute; Airbus DS GmbH
- country
Germany
- year
2010
- abstract
Payload scenarios of a robotic Lunar Cargo Lander (LCL) led to investigations concerning transportation and trial operations of an ISRU pilot plant for oxygen generation integrated into the Cargo Lander injected by Ariane5 into Lunar Transfer Orbit. The LCL carries several Lunar rovers and is equipped with davits for lowering the cargo to the lunar surface. The paper delineates the concept of the mission and corresponding system design and operations. The pilot plant mission and the real plant will need rover applications that are eminently in need of support by Artificial Intelligence (AI) as they comprise prospecting, ore identification, analysis, collection, preprocessing and transportation, all of which are processes that have to be accomplished in the confusing and unpredictable lunar terrain. Mining is followed by oxygen plant operation with overall process control including ore feeding, vacuum pyrolysis, slag removal and utilization, oxygen collection and processing along with control and optimization of solar power supply. The degree of autonomy of all these operations has a profound impact on overall costs. As a consequence, they will have to rely heavily on applications of artificial intelligence and knowledge based decisions for plan evaluation, goal reallocation, risk assessment and response to problems, all with as little human intervention as possible. The paper describes these and presents results of investigations about the details of the elements of a conceptual architecture of operations based on AI, cognition and robotics. Operations are carried out by robots that may or may not have android features. AI and cognition algorithms, such as rule and model based reasoning or evaluation of affordance and skill-oriented control, support the development of robotic platforms that are capable of perceiving an unstructured and dynamically changing environment. They are also capable of reasoning and predicting environmental changes and finally to synthesize appropriate actions, thus removing or mitigating the requirement for cost intensive ground support. While examples in the paper specifically refer to the operations of surface assets as required for solar pyrolysis of lunar surface material (Regolith), the results are in general applicable to all kinds of ISRU operations, including collection and electrolysis of lunar water.