The commercialisation of outer space by the US and Soviet Union in the late twentieth century brought revolutionary services, such as satellite telecommunication, remote sensing, and global navigation systems (Zhao, 2018). However, in the next century, access to outer space became available to companies seeking business opportunities in Earth’s orbit (Zhao, 2018; Beck, 2019). Whereas remote sensing satellites enabled the collection of geospatial data to study the Earth (Zhao, 2018), constellations of small satellites now provide access to the Internet in the most remote areas of the planet (Starlink.com, 2022). Consequently, the growing demand for real-time monitoring and analysis systems requires reliable, fast, and affordable satellite launch services (Rocket Lab, 2021b). However, what is the role of Big Data in the commercialisation of outer space, and if any, how does it influence decisions made by the stakeholders at those companies? These and other questions are explored in this work by analysing Rocket Lab, a small satellite launch company. The main themes identified in this work include the application of Big Data in system design, mission control and planning, manufacturing, and business management. This work discusses how the company incorporated Big Data into their business model, strategic decision-making, and whether Big Data aligns with the organisational goals or not. Finally, this work discusses professional and ethical issues related to the activities in outer space enabled by the use of Big Data.
Application of Big Data
Founded in 2006, Rocket Lab is a company specialising in building small rockets and spacecraft (Beck, 2019). The company's goal is to enable easier access to outer space, by providing reliable and frequent satellite launch services to its customers (Rocket Lab, 2021a). Throughout the production, testing, and launch activities, the company is generating a large quantity of real-time data (SDS, no date). That being said, Rocket Lab is handling Big Data, or BD, that can be defined as a collection of massive and complex datasets brought together from different sources (Al-Badi et al., 2018). The role of Data Analysts at Rocket Lab is to apply machine learning and statistical methods to extract meaning from BD (SDS, no date). This demonstrates that the company needs data-driven, decision-making capabilities, a powerful approach to harnessing data to make better-informed decisions (McAffee et al., 2012).
Big Data in System Design
Growing demand for satellite launch services is posing a challenge, as Rocket Lab tries to lower the cost and reduce the time needed to launch a satellite into orbit (Darley and Beck, 2021). Therefore, organisations turn to digital technologies to lower costs and improve operational efficiency and productivity (Manyika et al., 2011). For example, a Computer-Aided Design is a tool that combines data from different sources to enable system designers to create more efficient components. Since the space industry generates a vast amount of telemetric and sensory data (Buchanan et al., 2015), Data Analysts can analyse the data from real or simulated test flights (Badea et al., 2018) to monitor the effectiveness of parts during the launch, and address any issues related to inefficient design (Brown et al., 2011). Moreover, the company analyses data from the recovery flight computers and Guidance, Navigation, and Control (GNC) systems to detect patterns and anomalies, in order to transform them into actionable knowledge (Darley and Beck, 2021). Furthermore, data from tests conducted inside aerodynamic tunnels and test flights enables system designers to tune up the systems responsible for the stability and safety of the rocket during the flight (Buchanan et al., 2015).
Big Data in Planning and Mission Control
The company’s Mission Control Centre is a facility that provides launch vehicle monitoring and communication capabilities (Tulp and Beck, 2021). Similar to air traffic control systems (Badea et al., 2018), the mission control’s computers receive data streamed down from the launch vehicle sensors, such as pressure, temperature, and speed, that the company’s data analysis framework analyse and visualise (SDS, no date). As shown in Figure 1, the visualisation of data provides the mission control personnel with a tool to accurately measure and monitor the operating conditions of the launch vehicle, e.g., fuel supply, altitude, or propulsion systems (Tulp and Beck, 2021).
The mission planners also collect geolocation data to track the location of the rocket booster as it splashdown into the ocean after the second stage separation. The geolocation data enables the company to locate and retrieve the booster for analysis and reuse (Darley and Beck, 2021). Moreover, Rocket Lab collects data about the location of the spacecraft in relation to its orbit around Earth (Darley and Beck, 2021). The ground facilities, such as launch platform, fuel supply, or water coolant, also generate data, that mission planners analyse to “improve the integration, command, and technical support capability” of future missions (Wei Dong et al., 2019, pp.3). Besides scheduling preemptive measures to lower the risk of failure (Brown et al., 2011), Rocket Lab uses data to configure their launch vehicles to meet the mission-specific requirements (Darley and Beck, 2021), e.g., orbital entry trajectory and fuel requirements.
Big Data in Manufacturing
In manufacturing, any process that can be monitored are generating data (Brust, 2013). This data can be used to diagnose problems and act on them (Manyika et al., 2011). However, companies that keep data, can use it to build predictive models, for example, to predict when components are due to be replaced (Brust, 2013). Manufacturers can also collect a massive amount of sensory data, for example, through Radio Frequency Identification (RFID) devices that track the movement and location of larger items, smaller articles, or employees working in the factory (Manyika et al., 2011). This variety of data can be used to map out the most efficient assembly line configuration on the factory floor (Badea, et al., 2018). The analysis of data from manufacturing activities helps Rocket Lab to quickly identify problems, reduce costs (Beck, 2019), or improve manufacturing processes and efficiency as it has been done by other companies (Brown et al., 2011).
Big Data in Business Management
An increasing number of companies see BD as an important part of the company’s business model, that has been reported to improve productivity, efficiency, and growth (Schroeder, 2016). Data analysts scrutinise BD from different organisational functions to produce insights (Sivarajah et al., 2016) that Rocket Lab can use to evaluate its progress towards organisational goals. The literature review shows that companies use specialised tools, such as Business Intelligence, to analyse and visualise data (Manyika et al., 2011), or create self-regulating processes that automate many of the organisational functions (Brown et al., 2011). Data analysis can also inform business decisions, or point towards research and development of technologies that provide better value (Schroeder, 2016).
The literature review shows, that companies improve their financial performance either by using BD to forecast future outcomes (Schroeder, 2016), or by applying predictive analytics to decrease operational costs, improve sales, or create more efficient risk management (Manyika et al., 2011). Considering that rocket manufacturing represents 50% of the mission costs (Darley and Beck, 2021), Rocket Lab can apply inquisitive and predictive analytics to produce insights and guidance for the allocation of resources (Sivarajah et al., 2016). However, to maximise the application of BD, the management must be open-minded when insights from data contradict their own judgment (McAfee et al., 2012).
Challenges and Opportunities
Big Data requires efficient analytical capabilities that many companies struggle to achieve. Integration of data from different sources requires robust and reliable solutions to collect, store and analyse the data, as well as to develop effective analytical approaches for decision-making (Philips-Wren and Hoskisson, 2015). To illustrate, an average 6-hour flight of passenger aircraft generates 240 TB of data (Badea et al., 2018). In comparison, during a short test of a rocket engine, the monitoring systems generate petabytes of unstructured data that is difficult to store and analyse (Wei Dong et al., 2019). Although modern technologies, such as MapReduce or Hadoop address organisation and query issues (Sedkaui, 2018), data complexity, costs, and shortage of professionals with strong analytical skills remains a challenge (McAfee et al., 2012).
Big Data Challenges
The literature review shows, that data quality, duplication, and high volume, are three major causes of challenges faced by many companies (Al-Badi et al., 2018). For example, a decreasing cost of data storage enables companies to store more data (Brust, 2013). Consequently, companies are enticed to store data that in the long term, does not have any business value, is incomplete, or duplicated (Brust, 2013). Moreover, as more devices become interconnected in what is known as the Internet of Things (IoT), the velocity and variety of data are also becoming a challenge (Manyika et al., 2011). Moreover, there is a consensus among researchers that BD is a multi-disciplinary area that requires data analysts with multi-disciplinary skills (Datcu et al., 2020), and managers who understand how to make data-driven decisions (Brown et al., 2011; Allas et al., 2018). However, due to the heterogeneous and voluminous nature of data, BD management is also challenging (Ravat and Zhao, 2019). Furthermore, inaccurate, incomplete, biased, objective, or datasets without the context are likely to provide decision-makers with inconclusive evidence, lead to unjustified actions, or misguided decisions (Pagallo, 2017). Consequently, Rocket Lab could have a vast amount of data that does not have any value.
Big Data Opportunities
The literature review shows, that the opportunities offered by BD often outnumber the challenges. Data mining, knowledge extraction, or pattern recognition enable data scientists to 'make sense' of large datasets (Sedkaui, 2018). The increasing volume of data and computing power enabled significant progress in the development of Machine Learning (ML) and Deep Learning (DL), advanced analytical methods that brought significant breakthroughs in Artificial Intelligence (AI). AI has been loosely defined as the ability of machines to solve a range of problems without human intervention or detailed instructions (Allas et al., 2018). This technology provides capabilities, such as management of the satellite constellations, status monitoring, or onboard data analysis (Goodwill, 2021). However, the biggest opportunity for Rocket Lab is the AI’s capability to respond to different situations or objects (Goodwill, 2021). As the AI revolutionise other industries, Rocket Lab is likely to adopt this technology to optimise their manufacturing processes, or to develop autonomous GNC systems to land the next generation of Neutron rockets back on the launchpad, the company’s goals towards reusability (Beck, 2019).
Strategic and Operational Use of Big Data
By drawing from the examples shown in the literature and disclosed by Darley and Beck (2019), previous sections discussed areas where Rocket Lab applied BD. However, an increasing number of companies are using BD to support operations and strategic planning (Al-Badi et al., 2018).
Big Data Infrastructure
Previous studies show that companies use BD to improve their operational efficiency by capturing data that is relevant to their business objectives (Sivarajah et al., 2017). Since the space industry generates highly dimensional data that is exponentially growing in volume and variety (Wei Dong et al., 2019), an effective data strategy needs to describe what type of data must be collected, stored, analysed, and accessed (Sivatajah et al., 2017). Therefore, effective data infrastructure is needed to capture, store, and process a large amount of data to ensure that data captured by sensors and other devices is useful (Schroeder, 2016). However, conventional technologies, such as monolithic and silo-centric architectures are no longer suitable to meet the growing demand for access to data on-site or remotely (Manyika et al., 2011). For this reason, companies turn to data- centric architectures, such as Data Warehouses, Data Marts, and Cloud Computing platforms that offer economic and scalable data storage, processing, and computation services on demand (Manyika et al., 2011).
Big Data Analytics Methods
Data-driven companies use Data Analytics to get answers from the captured data (Schroeder, 2016). Those answers provide insights that Rocket Lab can use to support strategic planning and operations (Brown et al., 2011). An example of Data Analytics software is Business Intelligence (BI), a tool that enables the creation of reports, data analysis, and visualisations (Manyika et al., 2011). For example, the aerospace industry uses predictive and anticipative analytics to convert data into information that is used to save time and improve the efficiency of maintenance activities (Badea et al., 2018). Moreover, companies use Data Analytics methods, such as A/B testing to find the difference between the control and treatment groups (Manyika et al., 2011). Furthermore, companies use ETL methods to extract data from external sources (Manyika et al., 2011). AI has been deployed to automate manufacturing processes (Allas et al., 2011), and data visualisations, such as history diagrams, are being used to analyse the software engineering lifecycle (Manyika et al., 2011). Rocket Lab can apply similar methods to improve operational efficiency, quality, and lower operating costs. A/B testing can be used to determine changes in the rocket design. Whereas data visualisation has been found useful in mission control and planning, as discussed in section 2.2.
Insights from Big Data
The literature review shows, that companies analyse internal and external data from organisational functions and external partners to create insights (Manyika et al., 2011). Insights amplify risks and reveal patterns that otherwise could be left unnoticed. Insights help with understanding how organisational functions, such as supply chains, manufacturing, and distribution channels perform in comparison to the company’s estimates (Manyika et al., 2011), and provide a basis for better- informed decisions (Sravanthi, 2015). Furthermore, companies use BD to carry out controlled experiments of their investment decisions and to guide operational changes before their implementation (Brown et al., 2011). For example, in case of a rocket malfunction, Rocket Lab uses the insights to find the cause of the problem (Beck, 2019). Furthermore, space agencies use data from launch missions and daily operations in the management of space launch sites to assist the launch facility operations, such as launchpad preparation, rocket refueling, and safety checks before the launch (Wei Dong et al., 2019).
The Bottom Line
Although previous sections discussed numerous benefits of BD, the literature review also shows opposing arguments. Numerous studies highlighted that previous experience and own judgment is important when making data-driven decisions (Manyika et al., 2011; Hoffman, 2019) because results from data could be biased or based on incomplete data (Kennet and Redman, 2019). Therefore, Rocket Lab should be cautious when using BD as a tool to support its decisions. Moreover, previous studies also highlight that transition into the data-driven culture within well- established companies, with a long history, culture and certain mindset developed over the years (McAfee et al., 2012) is causing bottlenecks in the implementation of BD. Although those bottlenecks are unlikely to face Rocket Lab as it is a young company (Beck, 2019), their strategic and operational use of BD can be constrained by the lack of data analysts and managers with strong analytical skills, reflected on the company’s careers pages (Rocket Lab, 2022).
Professional and Ethical Requirements
Previous studies mostly defined professional and ethical issues as concerns surrounding BD and computer algorithms (Al-Badi et al., 2018; Hoffman, 2019; Kennet and Redman, 2019). The lack or poor implementation of data governance frameworks that influence the operational and strategic decisions made within companies has been argued to be an important professional issue (Al-Badi et al., 2018). Whereas certain studies focused on issues such as data storage and data policies (Al-Badi et al., 2018), the quality of data has been strongly identified as the largest and most impactful implication on decision-making in any organisation (Kennet and Redman, 2019). This claim is reasonable because poor data quality has been found to be “the emerging problem of bias for computerised decision-making” (Hoffman, 2019, pp. 901).
Professional Issues
Previous research and case studies showed that governmental space agencies experienced professional issues in the past. The maiden launch of the Ariane 5 rocket in 1996 or a failure of Mars Climate Orbiter in 1998, demonstrate examples of how computer algorithms, and the quality of data they process, decide the mission's success or failure. The erroneous software that caused the explosion of the Ariane 5 rocket resulted in the loss of the spacecraft soon after the launch. The failure to convert imperial units into metric, resulted in the loss of spacecraft as it entered the Martian orbit too close to the atmosphere, minutes before reaching its destination (Leveson, 2004). Those examples illustrate professional issues related to poor communication, poor data governance, and software quality measures that Rocket Lab needs to consider in their organisation.
Legal and Ethical Issues
From the ethical perspective, Livingston (2000) argued that the initial problem is to decide whose ethical standards to accept. Firstly, Rocket Lab operates in two countries, and secondly, the company offers services to international customers (Beck, 2021). Livingston (2000) argued that ethical values differ between people, leaders, and countries. This places Rocket Lab in a position where management should evaluate their customers’ motives, e.g., the purpose of data collected by their satellites and the impact on society and the environment. Moreover, Hoffman (2019) argued that BD and algorithmic decision-making is creating bias and unfairness within the organisation, for example, the use of algorithms in the recruitment process has been argued by O’Neil (2017) to put certain individuals in a disadvantageous position. Pagallo (2016) and Hoffman (2019) called it ‘algorithmic discrimination,’ as decisions are made based on biased or incomplete data, consequently resulting in unfair outcomes that hinder individuals’ efforts to get employed.
Whereas Data Protection Act 2018 ensures the privacy of Internet users, the Outer Space Treaty 1967 regulates the use of outer space, the Moon, and other celestial bodies. However, these are often falling short or fail to keep up with new technologies (Zhao, 2018), therefore, introducing further legal, ethical and professional dilemmas encompassing aspects such as accountability and moral responsibility for action taken by Rocket Lab now, and their short- and long-term consequences in the future.
Conclusion
In conclusion, this report explored literature related to the application of Big Data (BD) in commercialisation of outer space, its challenges, opportunities, and the strategic and operational use of BD. The work discussed how Rocket Lab uses BD and what else could be done to support their decision-making. The report also explored areas that should be considered, such as lack of trained personnel, and risks associated with a poor quality of data. However, due to the limited availability of case studies investigating the role of Big Data within Rocket Lab, some examples from other industries were carefully selected and translated into the context of Rocket Lab. Therefore, future research should consider exploring the role of BD in commercialisation of outer space in more depth, especially when discussing BD in the context of rocket launch companies. Interviews, email correspondence, or annual reports could provide a useful source of information. Nevertheless, this work provides a good starting point for discussion and future research.
References
Al-Badi, A. Tarhini, A. and Khan A.I., 2018. “Exploring Big Data Governance Frameworks.” In Procedia Computer Science, 141, pp.271-277.
Allas, T., Bughin, J., Chui, M, et al. 2018. “Artificial intelligence is getting ready for business, but are businesses ready for AI?” In Analytics comes of age. McKinsey Analytics.
Brust, A. (2013) Andrew Brust Processing Big Data [Audio] Available at: https:// stream.redcircle.com/episodes/3c482002-d20f-455f-b80f-5b9395f0243c/stream.mp3 [Accessed: 1 March 2022].
Beck, P. (2019) Small rockets are the next space revolution | Peter Beck [Video] Available at: https://www.youtube.com/watch?v=DhnBn_c9f8Q [Accessed: 17 February 2022]. Buchanan, G., Wright, D. Hann, C., et al., 2015. “The Development of Rocketry Capability in New Zealand - World Record Rocket and First of Its Kind Rocketry Course.” In Aerospace, (2)1, pp.91-117.
Brown, B., Chui, M. and Manyika, J., 2011. “Are you ready for the era of ‘big data’” McKinsey Quarterly, 4(1), pp. 24-35.
Darley, M. and Beck, P., 2021. “Return to sender: Lessons learned from rocket lab’s first recovery mission.” In 2021 Small Satellite Conference, pp.1-7.
Dong, W., Xiao, L., Niu, S., et al. 2019. “Application Research of Big Data for Launch Support System at Space Launch Site.” In Proceedings of the 3rd International Conference on Computer Science and Application Engineering (CSAE 2019), ACM, New York, NY, USA, Article 23, pp.1–6. doi:https://doi.org/10.1145/3331453.3360973.
Datcu, M., Le Moigne, J., Loekken, S., et al. 2020. “Special Issue of Big Data From Space.” In IEEE Transactions on Big Data, 6(3), pp.427-429.
Goodwill, J., Crum, G., MacKinnon, et al., 2021. “NASA SpaceCube Edge TPU SmallSat Card for Autonomous Operations and Onboard Science-Data Analysis.” In 35th Annual Small Satellite Conference, pp.1-13.
Hoffman, A.L., 2019. “Where fairness fails: data, algorithms, and the limits of anti discrimination discourse,” In Information, Communication & Society, 22(7), pp.900-915, DOI:10.1080/1369118X.2019.1573912.
Kennet, R. and Redman, T., 2019. The real work of data science. Hoboken, NJ: Wiley. Leveson, N.G. 2004. “Role of Software in Spacecraft Accidents.” In Journal of Spacecraft and Rockets. 41(4), pp. 564-575.
McAfee, A., Brynjolfsson, E., Davenport, T.H., et al., 2012. “Big data: the management revolution.” Harvard business review, 90(10), pp.60-68.
Badea, V.E., Zamfiroiu, A. and Boncea, R., 2018. “Big data in the aerospace industry.” In Informatica Economica, 22(1), pp.17-24. doi:10.12948/issn14531305/22.1.2018.02.
O'Neil, C., 2017. The era of blind faith in big data must end. [online] Available at: https:// www.ted.com/talks/cathy_o_neil_the_era_of_blind_faith_in_big_data_must_end?language=en [Accessed 03 March 2022].
Phillips-Wren, G. and Hoskisson, A., 2015. “An analytical journey towards big data” In Journal of Decision Systems, doi:10.1080/12460125.2015.994333.
Ravat, F. and Zhao, Y. 2019. “Data Lakes: Trends and Perspectives.” In International Conference on Database and Expert Systems Applications (DEXA 2019), Aug 2019, Linz, Austria, pp. 304-313.
Rocket Lab (2022) Rocket Lab to Expand Colorado Footprint with New Space Systems Complex. [Online] Available at: https://www.rocketlabusa.com/updates/rocket-lab-to-expand-colorado- footprint-with-new-space-systems-complex/ [Accessed: 3 March 2022].
Rocket Lab (2021a) Rocket Lab Successfully Launches U.S. Space Force Mission. [Online] Available at: https://s28.q4cdn.com/737637457/files/doc_news/2021/07/Media-Release_Rocket- Lab-Successfully-Launches-U.pdf [Accessed: 2 March 2022].
Rocket Lab (2021b) Rocket Lab Inks Deal to Launch Five Missions for BlackSky Constellation. [Online] Available at: https://s28.q4cdn.com/737637457/files/doc_news/2021/03/Media- Release_Rocket-Lab-Inks-Deal-to-Launch-Five-Missions-for-BlackSky-Constellation_23- March-2021.pdf [Accessed: 2 March 2021].
Schroeder, R. 2016. “Big data business models: Challenges and opportunities.” In Cogent Social Sciences, 2(1), doi:10.1080/23311886.2016.1166924.
SDS (no date) Senior Software Engineer: Operational Data. [Online] Available at: https:// www.space-defence-security-jobs.com/satcom/6670/senior-software-engineer-operational-data/ [Accessed: 26 February 2022].
Starlink.com (2022) World’s Most Advanced Broadband Internet System. [Online] Available at: https://www.starlink.com/satellites [Accessed: 4 March 2022].
Zhao, Y., 2018. “Space Commercialization and the Development of Space Law.” Oxford Research Encyclopedia of Planetary Science. Retrieved 28 Feb. 2022, from https://oxfordre.com/ planetaryscience/view/10.1093/acrefore/9780190647926.001.0001/acrefore-9780190647926- e-42.
