Mission handbook / 04 / integration

Engineering: Building a Working Mission

The best individual component is not necessarily the best choice for the whole spacecraft.

Write the question before choosing the hardware

Systems engineering is the practice of making the parts of a project work together to meet a defined purpose. For a space mission, that includes hardware, software, people, procedures and ground infrastructure. It starts with what users need to know or accomplish, not with an appealing spacecraft illustration.

Our example mission asks how vegetation changes after a dry season. That question needs refinement. Which area matters? How often must observations arrive? What uncertainty is acceptable? Who will use the result? A requirement converts a need into a statement that can be evaluated. “Take good pictures” is an aspiration; a defined coverage and measurement-quality target can guide a design.

Requirements must be consistent with resources. An optical sensor cannot guarantee an unobstructed view through cloud. Promising frequent delivery may require additional spacecraft, another sensing method or a more modest definition of service.

Balance electricity over an entire orbit

Solar arrays generate electricity when illuminated, while batteries store energy for periods when generation is insufficient. Power is the rate at which energy is transferred. Energy is the accumulated quantity. A system can meet a short peak-power demand yet still lack enough stored energy to complete a long activity.

A power budget lists expected consumption across operating modes. It must consider conversion losses, temperature, equipment degradation and uncertainty. A high-rate downlink may require more power than quiet monitoring, and charging a battery competes with other demands during sunlit periods.

Worked example: distinguish watts from watt-hours

A hypothetical device drawing 20 watts for half an hour uses 10 watt-hours. A real battery must provide more than this nominal calculation suggests because usable capacity, conversion losses, temperature and protective limits matter. This example explains units; it is not a battery-sizing recommendation.

Temperature control is energy management

In a vacuum, a spacecraft cannot cool itself by moving surrounding air as a desktop fan does. Heat moves within the spacecraft by conduction and between surfaces and the environment by radiation. Incoming sunlight, reflected light, Earth's thermal emission and internal electronics all contribute to the thermal balance.

Insulation reduces unwanted heat exchange; radiators provide surfaces intended to reject heat. Heaters can keep batteries or instruments within acceptable limits. “Space is cold” is therefore a poor design rule. A sunlit component can overheat while a shaded component becomes too cold.

Shallow-focus view of circuit-board components and a cable connector on an electronics laboratory bench
Electronics detail: electrical connections, heat paths and mechanical mounting all affect how an assembly behaves beyond the laboratory.

Know where the instrument is looking

Attitude means orientation, not orbital position. A spacecraft can be in the right place but looking in the wrong direction. Sensors such as star trackers, Sun sensors and gyroscopes help estimate orientation or rotation. Actuators such as reaction wheels, magnetic torque rods and thrusters change or maintain it.

Pointing requirements come from the payload and communications system. A camera may need stable viewing during an exposure, while a narrow-beam antenna needs alignment with a receiving station. These activities can compete, so a schedule may separate imaging, downlink and energy-recovery periods.

Onboard software coordinates these modes and responds to faults. A watchdog can help detect stalled execution; safe-mode logic can reduce activity and protect basic functions. Redundancy can improve resilience, but duplicate equipment is not automatically independent if both copies share the same vulnerable power line or software defect.

Turn measurements into a defensible map

An optical detector records a response to arriving electromagnetic energy. Calibration relates that response to physical quantities. Geolocation assigns measurements to positions on Earth. Atmospheric correction can help estimate surface properties by accounting for effects between the surface and sensor. Each step relies on models and introduces uncertainty.

Vegetation studies often compare red and near-infrared observations because healthy vegetation interacts differently with those wavelength regions. A vegetation index is a mathematical summary of selected measurements, not a direct universal reading of plant health. Soil background, species, growth stage, atmosphere and viewing geometry can influence interpretation.

Landsat, operated through NASA and the United States Geological Survey, and the Copernicus Sentinel-2 mission provide examples of multispectral observation. Sentinel-1 uses radar, an active system that transmits energy and measures the return. Radar can observe through many clouds, but interpreting its signal requires attention to surface roughness, moisture and geometry.

Spatial resolution describes the ground detail represented by a measurement; it is not a promise that every object of that size can be identified. A map also needs validation against independent information. Field observations and carefully selected reference data help assess whether the product supports its intended decision.

Verify the parts, validate the purpose

Verification asks whether the system meets its specified requirements. Validation asks whether it meets the actual user need in its intended context. Passing an electrical test does not prove a vegetation product answers the agricultural question that justified the mission.

Teams use interface documents, test procedures and configuration records to keep evidence connected to the hardware and software actually flown. Operational rehearsals exercise the ground team as well as the spacecraft. End-of-life actions, data documentation and long-term product access also belong in the mission plan.

The systems connection: collecting fewer, better-timed measurements can sometimes produce more useful results than maximizing raw data volume. Optimize for the mission's answer, not one impressive specification.

Read the underlying guidance

Unpack the remaining terms