For a constellation of agile Earth Observation Satellites (EOS), efficiently scheduling image acquisitions presents a complex decision-making challenge characterized by balancing a multitude of qualitative and quantitative preferences on the imaging requests while considering a high number of operational and temporal constraints. Current research predominantly focuses on the scheduling aspect, often neglecting the fuzzy multi-objective nature of the problem and the near real-time computation requirement. This study proposes an innovative three-stage solution method. Initially, an a priori multi-criteria scoring approach, based on the ELECTRE-III method’s fuzzy pairwise evaluation, is employed to value each potential imaging attempt, addressing the gap in comprehensive pre-scheduling valuation. The problem is then redefined as a Longest Path Problem in a Directed Acyclic Graph with Interdependent and Allowed Nodes (DAG-IAN). This re-conceptualization accommodates unique multi-satellite operational needs and imaging techniques such as stereo and strip acquisitions. We introduce the Extended Longest Path Algorithm (ELPA) for this purpose, which emerges as a novel solution mechanism. The final stage is a decision support system designed to guide decision-makers through the satellite operation’s intricate trade-offs, facilitating iterative enhancements and a deeper understanding of the conflicting objectives through a weight space analysis and a significance test. Our approach not only demonstrates high adaptability and explainability but also shows computationally efficient performance. In smaller problem scenarios, the ELPA closely approximates exact methods while significantly outperforming other approaches in large-scale applications. The research advances state of the art by offering an intuitive, customizable, and scalable framework in the preference integration aspect of the Satellite Image Acquisition Scheduling Problem.

We present a comparison study of state-of-the-art classical optimization methods to a D-Wave 2000Q quantum annealer for the scheduling of agile Earth observation satellites. The problem is to acquire high-value images while obeying the attitude maneuvering constraint of the satellite. In order to investigate close to real-world problems, we created benchmark problems by simulating realistic scenarios. Our results show that a tuned quantum annealing approach can run faster when used to find the optimal solution than a classical exact solver for some of the problem instances. Moreover, we find that the solution quality of the quantum annealer is comparable to the heuristic method used operationally for small problem instances, but degrades rapidly due to the limited precision of the quantum annealer.

The multi-satellite image acquisition scheduling problem is traditionally seen as a complex optimization problem containing a generic objective function that represents the priority structure of the satellite operator. However, the majority of literature neglect the collective and contemporary effect of factors associated with the operational goal in the objective function, i.e., uncertainty in cloud cover, customer priority, image quality criteria, etc. Consequently, the focus of the article is to integrate a real-time scoring approach of imaging attempts that considers these aspects. This is accomplished in a multi-satellite planning environment, through the utilization of the multi-criteria decision making (MCDM) models, Elimination and Choice Expressing Reality (ELECTRE-III) and the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS), and the formulation of a binary linear programming model. The two scoring approaches belong to different model classes of MCDM, respectively an outranking approach and a distance to ideal point approach, and they are compared with a naive approach. Numerical experiments are conducted to validate the models and illustrate the importance of criteria neglected in previous studies. The results demonstrate the customized behaviour allowed by MCDM methods, especially the ELECTRE-III approach.

We present a comparison study of state-of-the-art classical optimisation methods to a D-Wave 2000Q quantum annealer for the planning of Earth observation missions. The problem is to acquire high value images while obeying the attitude manoeuvring constraint of the satellite. In order to investigate close to real-world problems, we created benchmark problems by simulating realistic scenarios. Our results show that a tuned quantum annealing approach can run faster than a classical exact solver for some of the problem instances. Moreover, we find that the solution quality of the quantum annealer is comparable to the heuristic method used operationally for small problem instances, but degrades rapidly due to the limited precision of the quantum annealer.

Optical Earth observation satellites acquire images worldwide , covering up to several million square kilometers every day. The complexity of scheduling acquisitions for such systems increases exponentially when considering the interoperabil-ity of several satellite constellations together with the uncertainties from weather forecasts. In order to deliver valid images to customers as fast as possible, it is crucial to acquire cloud-free images. Depending on weather forecasts, up to 50% of images acquired by operational satellites can be trashed due to excessive cloud covers, showing there is room for improvement. We propose an acquisition scheduling approach based on Deep Reinforcement Learning and experiment on a simplified environment. We find that it challenges classical methods relying on human-expert heuristic.

Mission planning for Earth Observation Satellite operators typically implies dynamically altering how requests from different customers are prioritised in order to meet expected deadlines. A request corresponds to a given area of interest to capture on Earth. This exercise is challenging for different reasons. First, satellites are limited by maneuvers and power consumption constraints resulting in a limited surface that can be covered at each orbit. Consequently, many requests are in competition with each other and so all of them cannot be treated at each orbit. When a request priority is boosted, it may incidentally penalise surrounding requests. Second, there are several uncertain factors such as weather (in particular cloud cover) and future incoming requests that can impact the completion progress of the requests. With order books of increasing size and the planned operations of a growing number of satellites in a close future, there is a clear need for a decision support method. In this paper, we investigate the potential of Evolutionary Algorithms (EAs) for this problem and propose several approaches to optimise request priorities based on Local Search and Population-Based Incremental Learning (PBIL). Using a certified algorithm to decode request priorities into satellite actions and a satellite simulator developed by Airbus, we are able to realistically evaluate the potential of these methods and benchmark them against operator baselines. Experiments on several scenarios and order books show that EAs can outperform baselines and significantly improve operations both in terms of delay reduction and successful image capture. While black-box approaches yield significant improvements over the baseline when there are many delayed requests in the order books, introducing domain knowledge is required to handle cases with fewer delayed requests.