A satellite passes over an area worth imaging. The geometry is right, the instrument is available and the observation would be valuable. Yet taking the image consumes energy and onboard storage. The spacecraft may soon need to transmit data to a ground station, change its pointing direction or perform another activity that competes for the same resources. The decision is not just whether the picture can be taken. It is what taking it makes harder to do afterward.
Satellite mission planning is a scheduling problem shaped by orbital motion and spacecraft limits. Unlike appointments in an ordinary calendar, many activity windows are determined by geometry. Some opportunities occur only when an instrument is oriented appropriately; others depend on lighting or communication visibility. A plan must respect both these windows and the resources that carry across them.
In a useful model, energy state, data storage and activity timing are part of the decision, not checks added after a schedule has been drawn.
The observation that fills the memory
Imagine a spacecraft with enough free storage for two ordinary images or one unusually large one. The large image has high scientific value, but taking it would leave no room for another observation until a downlink becomes possible. If a ground-station contact occurs soon, the choice may be easy. If the next contact is much later, preserving space could allow several important observations that would otherwise be lost. A simple count of collected images or individual request priorities does not reveal the best plan.
Energy creates a similar connection through time. An activity may be feasible now but leave too little available power for a required later maneuver or instrument operation. In a useful model, energy state, data storage and activity timing are part of the decision, not checks added after a schedule has been drawn.
Priorities aren't a single queue
Missions can involve routine observations, urgent requests, calibration, maintenance and communication sessions. Some activities must precede others. Some are optional but valuable, and some cannot be postponed. A planning model distinguishes those categories and describes what is being optimized: scientific return, coverage, timely delivery of data or another mission-specific objective. The weights or priority order should reflect the mission's actual policy.
The optimization ideas are related, but the problems should not be conflated.
New information may arrive after the schedule has been prepared. A ground-station contact can be lost, an activity can overrun, or a high-priority request can appear. Replanning then becomes a balance between using the remaining opportunities and preserving commands that have already been committed. Optimization can help expose those alternatives but does not replace flight rules, engineering reviews or operational authorization.
Mission planning is not sensor scheduling on the ground
The optimization ideas are related, but the problems should not be conflated. Ground-based space-observation scheduling allocates telescopes, radar or other sensors to observations of objects in orbit. Satellite mission planning controls the activities performed by a spacecraft, with constraints such as its own pointing, power, memory and communication opportunities. Both involve fleeting windows; their physical models and decision responsibilities differ.
This distinction is why one universal scheduling template would be a poor starting point. The algorithm should exploit structure that is actually present in the mission model. Read observation scheduling for the ground-sensor perspective, resource-constrained project scheduling for the common mathematical foundation and Custom Optimization for integrated optimization and problems unique to your organization.