5 ways aerospace product development solves real-time challenges in space

The space industry is among the most innovative in the world.

Propelled by communications, exploration, and R&D possibilities, it is estimated that the global space economy could be worth $1 trillion by 2030 (McKinsey & Company, 2023).

However, space is an incredibly challenging environment—for life and technology alike—and the potential costs of failure are exceptionally high. Teams involved in aerospace development must work hard to mitigate these risks.

In this blog, we look at how aerospace product development has helped overcome these challenges.

Challenge 1: Detecting and isolating faults

Challenge 2: Prognostics and health management

Challenge 3: Slow communication

Challenge  4: Power management

Challenge 5: Bugs

No room for error

Challenge 1: Detecting and isolating faults

From radiation exposure and extreme temperatures to micrometeoroid impacts and manufactured debris, the harsh environment of space presents significant risks to spacecraft and their often delicate systems. If left unchecked, such hazards have the potential to cause component failures, leading to malfunctions that could jeopardise the success of a mission.

Cosmic rays pose a particular threat to space computers as they can physically change the state of a floating gate transistor, causing a 0 to become a 1 or vice versa. This is known as a bit-flip. the impossible outcome of a Belgian election.

Fault Detection, Isolation and Recovery (FDIR) is widely applied in space engineering to safeguard against these threats—from individual components to the entire craft. A type of control engineering, FDIR monitors sensor readings, actuator performance, and system behaviour to identify anomalies. When one is detected, it pinpoints the root cause and initiates corrective actions.

Efficiently diagnosing and correcting faults reduces downtime and increases system availability.

Challenge 2: Prognostics and health management

Not all faults are unpredictable. Spacecraft rely on a complex collection of components which have a finite lifespan. In space, where corrective or reactive maintenance may not be feasible, predicting component failures is essential for minimising downtime and ensuring mission continuity.

Historically, conditional maintenance would be used to predict component failure. This could be as simple as observing imminent problems (burning smell, excessive heat, increased running noise, and so on) or as complex as conducting vibration analysis using Control Charts. However, such methods are labour-intensive and not especially accurate.

Advances in machine learning have made predictive maintenance (PdM) possible. It employs historical data, multiple sensor readings, and operating conditions to form a multivariate prediction model which will estimate the remaining useful life of critical components. The prediction model will become more accurate over time as it acquires more data. It has proven highly cost-effective in manufacturing.

However, the aerospace industry is currently looking ahead to prescriptive maintenance (RxM), which takes this one step further. As well as predicting when components need replacing, RxM can provide proactive guidance for preventing failures and improving performance.

A 2023 research paper published by NASA cites the possibility of a 30% reduction in maintenance costs using RxM. However, it also identifies five key challenges to its implementation in aerospace: the complexity of prediction; validation, safety assurance, and regulatory challenges; cost of adoption; difficulty in quantifying impact and informing decisions; and data availability, quality, and ownership challenges.

Challenge 3: Slow communication

In the 60s, S-band radio waves allowed the world to watch the moon landing. While technology has come a long way since then, radio communication remains spacecraft’s most widely used form of communication. For communicating with devices that are relatively close to Earth, it’s pretty reliable tech, primarily limited by the latency of relaying information to a transmitter with line-of-sight to the intended receiver.

However, radio signals diffuse as they travel over large distances. The further they are broadcast, the more power we require to broadcast data, and the larger the antennae or array needed to receive it.

While Starlink satellites can provide download speeds of up to 220Mbps, the Mars Rover, 200 million kilometres away, can only achieve up to 2Mbps. Voyager 1, currently more than 23 billion kilometres from Earth, returns data to us in real-time at a speed of 160 bits per second.

This is a challenge for spacecraft as we explore further from Earth. Compounding the issue is that space on the airwaves is limited, and demand for space communications is increasing.

Aerospace may have developed an answer: lasers. Bursts of focused, near-infrared light can achieve data transfer rates—up to 100 times faster than radio—and should retain their signal strength over long distances. NASA is currently testing the technology across multiple missions. In November 2023, its Deep Space Optical Communications (DSOC) experiment successfully beamed high bandwidth test data back to Earth from 16 million kilometres away.

Challenge  4: Power management

While the costs have reduced significantly over the past decade, putting technology in space isn’t cheap. Every milligram counts, and system designs must be kept as lean as possible. This includes power sources and storage.

Since spacecraft rely on limited power sources—solar panels paired with Li-ion batteries, or nuclear power—effective power management is necessary to ensure critical systems have power when needed.

Well-designed power system management is necessary to control the distribution and storage of power. This software optimises utilisation, minimises loss, and provides backup power if the primary source fails. However, power management software must be complemented by efficient hardware design.

As our embedded power management blog post explains: “Balance is vital to ensure that the device is both operational and efficient.”

Challenge 5: Bugs

Significant advances in AI and automation are paving the way to an exciting new era of embedded software engineering for space innovation and discovery. However, such advances come with their own risks.

In 2019, Boeing’s partially reusable crew transport craft, the Boeing CST-100 Starliner, failed to dock with the International Space Station (ISS). A coding error caused the craft to mistakenly use the time settings of the mission’s Atlas V launch vehicle, thereby messing up timings for critical stages of the mission. Insufficient testing meant nobody had simulated a whole mission from launch to docking and this critical bug went unnoticed.

This was certainly not the first time clocks caused significant software problems.

Consequent software checks revealed a further bug that, if left unaddressed, would have caused Starliner to fire its thrusters at the wrong time, resulting in uncontrolled motions that could have caused a collision between the service and crew modules. Had this happened, the crew module’s heat shield could have been damaged, or the module could have tumbled dangerously as it returned to the surface.

An increasing reliance on automated systems must be balanced with the highest software development standards.

No room for error

Space presents many unique challenges for aerospace product design and development. The costs are high, the margins are tight, and a single failure could prove catastrophic. It is vital that every single element of space technology—materials, components, and software—are designed, constructed, and tested to the highest standards.

The “final frontier” is no longer the exclusive domain of nationally-funded organisations and the old guard of the aerospace industry. At a time when even some startups have the means to put devices into space, it’s more important than ever that safety and quality are not overlooked.