SpaceX: Enabling Space Exploration through Data and Analytics
SpaceX has disrupted the space industry by making launches more accessible – data has been key to their success
SpaceX has disrupted the space industry…
When you send a rocket to space, that rocket is exposed to different forces: gravity, changes of pressure, changes of temperature, aerodynamics, and acoustics. These forces produce responses in the structures of the spacecraft, following Newton’s second law of motion (F = ma), which make it hard to propel and navigate. These responses can also be detrimental to the systems and astronauts onboard.
Given this harsh space environment, the development of spacecraft is complex, requiring various design and analysis cycles to meet the required technical specifications for mission success. The process involves lead designers who develop the initial design using computer-aided design (CAD) software. This design is then sent to various groups of technical experts, which conduct a series of analyses and tests. These analyses and tests produce data and analytics that assess the performance of the design against each technical team’s specified requirements. This process is repeated until the spacecraft’s design meets all desired performance indicators. See Exhibit 1 below for the full spacecraft development cycle.
Figure 1: Spacecraft Development Cycle
Through this development process, SpaceX has innovated in a variety of areas:
Data for Development
SpaceX reduced the cost of access to space by a factor of 10. This cost reduction is mainly driven by its investments in specialized software and digital platforms to increase efficiency of the above mentioned process through improved data management. These CAD and finite element analysis software, for example, store rocket assemblies and databases that are shared across various teams of experts through a centralized repository. The software is fast, which encourage engineers to perform rapid iterations of both its virtual prototypes. The centralized nature of databases also promotes communication and collaboration of SpaceX’s teams, thus removing silos.
The abundance of data through the development process also allows SpaceX to identify bottlenecks and assign resources to specific efforts as necessary. Data also allows real-time monitoring of the development cycle, something that was not easily accessible in the past given the legacy of the industry which utilized analog systems and processes developed during the space race of the 1960s.
Finally, the data-heavy development process enables innovation of more sophisticated and advanced systems. As data analytics reveal the impact of multiple variables in the performance of a rocket, engineers are able to make better informed decisions about design specifications and constraints. SpaceX, as an example, has developed and flown various reusable Falcon 9 rockets (see Figure 2) thanks in part to its digital development approach.
SpaceX’s investment in digital platforms has resulted in 50% productivity improvement.
Figure 2: Reusable Falcon 9 rocket
Data for Manufacturing
Technicians on the SpaceX shop floor look at data and models to better understand a rocket’s inner workings through its manufacturing and assembly. This is particularly helpful for seeing the specs of internal systems, such as electrical wirings, thus improving efficiency. Data creates an intelligent factory setting, in which physical and virtual assets are utilized to drive better performance.
Data also enables the adoption of new manufacturing processes, such as 3D printing. These new manufacturing processes require big data inputs, and ultimately enable SpaceX to more vertically integrated, which is part of the company’s business model (SpaceX currently produce over 85% of its launch hardware in-house). This allows them to further reduce costs, thus helping them achieve their mandate.
Figure 3: Dragon Crew Operations
Data for Operations
SpaceX also utilizes digital twins in its operations, which are virtual representations of physical assets. These digital twins utilize virtual simulation and real-time data to rapidly evaluate and monitor systems and optimize its operations through the use of data analytics.
Digital twins enable operators in Mission Control to have a real digital replica of SpaceX’s Dragon vehicle (Figure 3) to monitor its status – trajectory, loads, propulsion systems, etc. – from data received from the hundreds of sensors integrated in the spacecraft. This data ultimately increases reliability and safety of SpaceX’s Dragon and other vehicles.
In conclusion, SpaceX has been able to maximize the impact of its data in the development, manufacturing, and operations of its spacecraft. This is the result of heavy investments in end-to-end development systems that enable the monitoring and deriving of insights from its development cycle – this improving efficiency and driving innovation. Additionally, data enables improvement of their manufacturing and assembly processes and the adoption of new techniques such as 3D printing. Finally, SpaceX has been successful at utilizing digital twins to track and monitor its systems while in orbit, thus improving safety and performance.
Student comments on SpaceX: Enabling Space Exploration through Data and Analytics
This is SO cool! Thank you for sharing this Juan Carlos.
1) I was under the impression that the cost reduction by a factor of 10 was due to reusing boosters, and due to NASA’s inherent organizational inefficiency in spreading it’s operations around several states for political reasons (spreading around federal dollars more equitably). How dose data play into that?
2) That digital twinning concept is SO COOL! Has SpaceX created simulator/models that are able to capture all the dynamic conditions involved in launches?
3) Would be thrilled to buy you a beer or coffee to discuss this stuff further!
Thanks again for an awesome post!
Thank you for commenting, Omar!
1) Yes, the reusable boosters drive costs down, however, their development was possible thanks to SpaceX’s innovative data handling and operations, coupled with agile processes and effective leadership which resulted in substantial productivity improvements. NASA, due to the inefficiencies and bureaucracy you mentioned, has been very slow at adopting digital technologies, which often result in lost opportunities or misuses of their data. SpaceX, on the other hand, adopted those digital systems almost since its inception, which cut substantially the development schedules, costs, and ultimately allow SpaceX to be more vertically integrated and agile. While NASA continues to develop Orion and SLS, SpaceX has flown Falcon 9, Falcon Heavy, Crew Dragon, Starship, etc. NASA recently launched a Digital Transformation initiative to address these issues and make better use of their data, hopefully they can catch up soon: https://www.nasa.gov/offices/oct/what-is-digital-transformation-and-why-is-nasa-doing-it.html
2) Yes, digital twins have become popular in the industry: https://www.youtube.com/watch?v=geS6wsUHmOw. There are obviously still some complexities those simulations are not able to capture (i.e. radiation is typically hard to model), but we are getting there…
3) I would be down for a beer or coffee any time!
Juan Carlos, I learned a lot from your article. Thank you for your post! One thing that has struck me about SpacEx in recent years is their push in the 5G arena. This is one company (along with Google, Amazon, etc…) that is outspending traditional aerospace and defense orgs in the 5G space. Typically, the military is the first to build out / get the next generation of advanced communication technology… not tech firms. So, talk about shift in market dynamics. In January of 2020, SpaceX successfully launched its fourth satellite compatible with 5G. It will be one of their thousands of global 5G network satellites SpaEx will launch in the coming five to ten years.
Thank you for your comment, Cristina. I agree with you, SpaceX’s work on the 5G arena is truly impressive. The company is currently building a satellite constellation, Starlink, to bring internet via satellite to the world including remote and isolated areas. SpaceX’s agile operations allows them to integrate the latest technologies, such as 5G, to their fleet of satellites. A lot of satellite developers fail to do this: by the time they launch their satellites into orbit, a lot of their technologies and systems are obsolete.
Thanks Juan Carlos! Would love to hear your thoughts on how SpaceX’ use of analytics compares to NASAs in class. Separately, I wonder if there are tradeoffs to these new innovations in regard to quality assurance and if that presents further strategic problems and bottlenecks in the development process.
I appreciate the comment, John. NASA unfortunately has silos, preventing engineers like me who are focused on a specific technical area, to access data and analytics from other technical groups. This prevents teams from making real-time decisions on the potential tradeoffs, which result in decelerations of the development process. The debate is up on whether these innovative technologies don’t sacrifice quality and safety, at the expense of expedience. There should be a balance.
Fascinating Juan Carlos, two things come to mind. 1) Would love to hear about the work SpaceX’s competitors, such as Blue Origin, are doing to get some sort of private company benchmark. 2) Is there any cooperation with NASA in the digital arena, as in some knowledge transfer for the public good?
Great comments, Francisco. 1) Blue, like SpaceX, has been successful at integrating these digital technologies into their operations; unlike the incumbents which integrated those technologies later. Blue, however, has a different business model. Their mantra “gradatim ferociter”, meaning “step by step ferociously”, symbolizes the company’s approach to space flight. They are less worried about rapid operations and more focused on building new technical capability. 2) There is a lot of cooperation between NASA and New Space companies such as technology transfers and licensing, access to data, and technical experts. Data, however, is not typically package nicely for private companies — they often have to mine that data or get it through the technical experts.