Edward F. Crawley’s research while affiliated with The American Institute of Aeronautics and Astronautics and other places

Conceptual State Analysis (CSA) is the process of defining, exploring, and reasoning about state attributes and state spaces in complex system architectures. Model-Based State Analysis enhances CSA with formal modeling and analysis methods. We propose a process for generating the state vector of a given system architecture, based on a graph representation of the architecture’s model in Object-Process Methodology. A robust graph data structure that represents the model is queried to produce the architecture’s state space. The process facilitates analysis, reasoning, and model revision based on improved understanding of state vectors and state spaces. State attribute refinements within the model allow for model validity, viability, and reusability as the architecture evolves, with multiple agents, attributes, and attribute state values. The CSA approach advocates and facilitates careful, dynamic, and interactive state space exploration in lieu of exhaustive upfront enumeration of state space permutations. The results can be fed into other analysis, simulation, or visualization tools. We demonstrate CSA on driver assistance technology.

Fuel selection is a strong driver of the mass fraction for many proposed lunar and Mars missions, but fuel technology trends have not been comprehensively evaluated for their impact on the system in literature. We evaluate the impact of fuel selection on overall lunar architectures. Our analysis shows that although hydrogen architectures have a higher wet mass cost, they provide more payload capacity to the lunar surface than non-hydrogen architectures given the same number of campaign launches. The Moon has been viewed as a stepping stone for future planetary exploration, so we evaluate both Mars and lunar architectures. We functionally decompose architectural decisions and compare key campaign decisions across 18 notable Mars architectural studies. The 18 landers are classified into four groups depending on which of the four the functional capabilities the lander performs, namely outbound transit, mars descent, mars ascent, and inbound transit. We find that there is no strong relationship between the Martian landers’ wet mass and the length of crewed Martian surface. Furthermore, fuel type selection did not have a clear trend with the aforementioned capabilities. The lack of similarities across Mars architectures suggests the reference studies had a wide range of depths of analysis along with an array of different methods. Furthermore, they were completed at various points in history, some with high political pressure.

We describe a model-based technology roadmap that explores the capability of human landing systems to deliver crew to and from the lunar surface for short-stay ‘sortie’ missions. This kind of mission is assumed to take place through the first five years of future human lunar exploration and during the following early stage of the lunar outpost initial operational capability. After that, the human landing system will be able to continue transporting people to and from the Moon but will need to undergo some modifications to be able to stay dormant at the outpost for longer periods of time, up to 180 days. In this paper we describe the technology roadmapping approach that has been adopted. We use an Object Process Model to describe the lunar human landing system and supporting infrastructure. We define relevant figures of merit to measure performance and cost associated with lunar landing system developments. We provide a benchmark of the state of the art in lunar landing systems through a comprehensive literature review, and describe a technical model to set credible performance targets for future landing systems. In our analysis, we adopt NASA as the reference customer. We assume a landing crew of 4 people, operating on the lunar surface for 7 days. We assume departure and return points as the Lunar Gateway station in near rectilinear halo orbit. We assume a payload mass delivered to the surface of 500 kg, with additional 250 kg allowance. Based on a tradespace modeling exercise, we explored alternative system architectures for lunar landing systems. The analysis allowed us to set targets for future development, which we organized in a technology roadmap timeline. The roadmap suggests an informed path forward for future development of lunar landing systems, setting credible technology targets that are validated by the described model-based approach. All suggestions in the roadmap are backed up by technical models, which are briefly overviewed in the paper.

2020 The MITRE Corporation. All Rights Reserved. Fuel selection is a strong driver of the mass fraction for many proposed lunar and Mars missions, but fuel technology trends have not been comprehensively evaluated for their impact on the system in literature. We evaluate the impact of fuel selection on overall lunar architectures. Our analysis shows that although hydrogen architectures have a higher wet mass cost, they provide more payload capacity to the lunar surface than non-hydrogen architectures given the same number of campaign launches. The Moon has been viewed as a stepping stone for future planetary exploration, so we evaluate both Mars and lunar architectures. We functionally decompose architectural decisions and compare key campaign decisions across 18 notable Mars architectural studies. The 18 landers are classified into four groups depending on which of the four the functional capabilities the lander performs, namely outbound transit, mars descent, mars ascent, and inbound transit. We find that there is no strong relationship between the Martian landers’ wet mass and the length of crewed Martian surface. Furthermore, fuel type selection did not have a clear trend with the aforementioned capabilities. The lack of similarities across Mars architectures suggests the reference studies had a wide range of depths of analysis along with an array of different methods. Furthermore, they were completed at various points in history, some with high political pressure.

Liquid Oxygen liquid Hydrogen (LOX/LH2) Mars architectures have historically been ruled out in favor of Liquid Oxygen liquid Methane (LOX/LCH4) architectures due to a high perceived cost and risk in LH2 cryo-management, which leaves a gap in literature and technology development in support of LH2 architectures. Carbon is not thought to be abundant on the Moon, while recent discoveries suggest that water is widely accessible on Mars, presenting an opportunity. The challenges in meeting this opportunity are long-term LH2 cryo-management and fuel in-situ resource utilization (ISRU). To investigate this opportunity, a LOX/LH2 Mars architecture is modeled with combinations of the following technologies: passive and active cryo-technologies, fuel cells to produce water and power from the otherwise wasted boiloff, and fuel and oxidizer ISRU. As these technologies are switched on and off in combinations, we compare the resulting mass that has to be sent to Earth's orbit (upmass), which is a proxy for campaign cost. With improved multi-layer insulation (MLI) and cryocoolers, we forecast upmass savings of 6%-10% compared to the LOX/LCH4 baseline. The addition of fuel cells only gave a marginal benefit of 0.7% upmass savings, but they add operational flexibility for power and life support. Additionally, we show how much more payload could be landed on Mars given the same upmass as the baseline, and we project an improvement from 22 mT in the LOX/ LCH4 baseline to 31 mT in a LOX/LH2 architecture with cryocoolers and MLI. We also analyze ISRU in the context of multiple crewed campaign cycles and project that the added mass due to Mars water ISRU infrastructure pays off after the second crewed mission. These results suggest a Mars LOX/LH2 architecture with improved MLI and cryocoolers will be competitive with the traditional LOX/LCH4 architecture, especially under longer campaign plans.

The study and benchmarking of Deep Reinforcement Learning (DRL) models has become a trend in many industries, including aerospace engineering and communications. Recent studies in these fields propose these kinds of models to address certain complex real-time decision-making problems in which classic approaches do not meet time requirements or fail to obtain optimal solutions. While the good performance of DRL models has been proved for specific use cases or scenarios, most studies do not discuss the compromises of such models. In this paper we explore the tradeoffs of different elements of DRL models and how they might impact the final performance. To that end, we choose the Frequency Plan Design (FPD) problem in the context of multibeam satellite constellations as our use case and propose a DRL model to address it. We identify six different core elements that have a major effect in its performance: the policy, the policy optimizer, the state, action, and reward representations, and the training environment. We analyze different alternatives for each of these elements and characterize their effect. We also use multiple environments to account for different scenarios in which we vary the dimensionality or make the environment non-stationary. Our findings show that DRL is a potential method to address the FPD problem in real operations, especially because of its speed in decision-making. However, no single DRL model is able to outperform the rest in all scenarios, and the best approach for each of the six core elements depends on the features of the operation environment. While we agree on the potential of DRL to solve future complex problems in the aerospace industry, we also reflect on the importance of designing appropriate models and training procedures, understanding the applicability of such models, and reporting the main performance tradeoffs.

A Digital Engineering Information Exchange supports an exchange of digital artifacts between system engineering entities (INCOSE). Such entities might include processes, models, and organizational elements associated with space missions design. Reducing complexity and errors, as well as improving efficiency are critical capabilities associated with a digital transformation of space missions design and delivery. In our work we propose an approach to manage a digital engineering information exchange through the DSM-based approach (Eppinger and Browning 2012). Applied to space systems architecture, the method allows keeping track of the information exchange throughout the product development. Such information includes the core entities and relationships of CubeSat's subsystems. This would integrate systems engineering (MBSE) approaches and PLM methods. In our paper we apply the proposed approach to a CubeSat mission design. One of the forms of utility of the proposed approach is the ability to represent the subsystems and their interfaces including objects/processes/states in one DSM/DMM representation. In our paper we demonstrate how such evaluation can be performed. Another utility of the proposed approach is that it facilitates a digital information flow through different product lifecycle stages. A proposed approach might serve as an effective method to reduce complexity associated with different ontologies in different design tools. Ultimately, it allows engaging digital tools in a concurrent engineering environment.

Poster session at the 22nd International DSM Conference