Cognitive Constellation Management Scheduling

We have built our existing Freedom Platform as a cloud based SAAS system with one main goal: simplicity. Every feature we add is driving towards a vision of reducing the complexity of ground system operations and access. nbsp;We propose a technical solution that removes much of the constellation planning daily grind from the customer so they can concentrate on exploiting their data. The end result is a reduced staff, less room for error, and optimized access to data.nbsp;ATLAS currently uses a generalized highly parallelizable, polynomial-time schedule algorithm called the Flex Scheduler for task deconfliction across multiple ground systems with many customers. The Flex Scheduler is built to support task requestsnbsp;and optimize deconfliction.nbsp;It is important to note that this approach is the only way to achieve higher utilization of ground systems due to its ability to ldquo;load balancerdquo; across many ground systems and generate a globally optimal utilization plan. Without giving the customer the ability to generalize their requests a scheduling system can only be a deconfliction algorithm, not a system wide optimization function. Where most scheduling systemsnbsp;are intended for one satellite constellation, the Flex Scheduler is federated across all ATLAS owned systems and considers all customers simultaneously. Federation of the data allows for deeper insights.nbsp;To improve the Freedom Platform and take it to the next level, we propose the following:System Wide Load Balancing: Define clear goals and allow the customer to adjust their importance, develop a task decomposition phase that generates task requests, and feed them into the lower lever Flex Scheduler.Time Reduction: Abstract the management into a simple to understand set of goals that the customer can adjust on a day to day basis.nbsp;Rapid Re-Scheduling via Machine to Machine: Define a concise, cloud hosted, API that allows the submission of task requests along with metadata / spacecraft state parameters.