Technology-Determined Organizational Structure

Discussion

By Lucas Lemos and Chris Giotitsas:

CubeSat technology greatly determines the organization of work. ESTCube-1 consisted of seven subsystems and two ground support systems (Table 1). The communication and interrelations between subsystems must be well-defined and effectively executed. Each subsystem has a team leader who coordinates the work tasks inside the subsystem team. Also, team leaders coordinate tasks among themselves. All subsys tem teams engage in discussing the main choices from the early stages of design. All issues are open for discussion, and all members are involved in making crucial decisions, like choosing a launch provider (Slavinskis, Pajusalu, et al., 2015). Team leaders use transversal groups created on the spot to deal with newfound coordination needs. In areas of limited impact or requiring specific expertise, the discussion and decision making are entrusted to task-force groups. Students are usually involved in developing more than one subsystem to gain practical multidisciplinary collaborative working skills.

ESTCube outreach efforts resulted in more than 15 aca demic journal articles (and several conference papers and posters) documenting the technology, the design, the results, and the know-how of the project. Payloads and subsystems and thoroughly described from functions to hardware, and from technical specifications to the justification of design choices. These academic publications appeared mostly dur ing condensed periods—especially right after the development of ESTCube-1, and we could split them into two areas. First, publications connected with the CubeSat technology (design and flight results), including areas like the main pay loads (Iakubivskyi et al., 2020; Lätt et al., 2014); the electri cal power system (Pajusalu et al., 2014); or the command and data handling subsystem (CDHS) and its firmware (Laizans et al., 2014; Sünter et al., 2016). Second, publica tions dealing with the project and the community (know how), including a compilation of lessons learnt (Slavinskis, Pajusalu, et al., 2015); the working processes and the funding (Kalnina et al., 2018); and the management of the com munity (Noorma et al., 2013; Slavinskis, Reinkubjas, et al., 2015). We can say that, in the case of ESTCube, technology determines their organizational model, and their academic outcomes reflect the impact of technology determinations over how they organize themselves. These academic publications follow the same modular structures (payloads, subsystems) of the CubeSat.

ESTCube uses a network of partners to develop and man ufacture CubeSat’s technology. Often, companies agree to provide services or discounts to support ESTCube. During the development of ESTCube-1, the preference was to use customized components off-the-shelf (Slavinskis, Pajusalu, et al., 2015). However, not all components are available in Estonia, and in some cases, they were shipped from the United Kingdom, Germany, or the United States. Hardware components are provided with different—primarily closed— licenses. Students usually utilize proprietary software pro vided by the universities to students and academic staff. This tendency is shifting as the advantages of working with open source software become more evident to ESTCube members.

The ESTCube-1 CubeSat incorporated open-source soft ware. The mission control system of ESTCube-1 used open source software “Hummingbird” codeveloped with an Estonian company (ESTCube, 2020). The open-source “FreeRTOS” operating system was used on the satellite’s main onboard computer (CDHS) and the camera, while the also open-source TinyOS operating system was used on the communication module. As part of the results and lessons learnt from ESTCube-1 experience, the use of an operating system that “provides most of the needed functionality, for example, a form of embedded Linux” (Slavinskis, Pajusalu, et al., 2015, p. 18) is recommended.

It is, however, during the development of ESTCube-2 that open-source software becomes widely used. The open source “tech stack” (a set of technologies used to build a single application) of ESTCube-2 developers consists of nearly 20 digital technologies including languages (Python, Go, Apache Groovy), databases (PostgreSQL, SQLite) and satellite specific software like Skyfield (see Table 2). Besides, the development setup (Microsoft Visual Studio Code), and the repositories (Git) used are open source. To provide quality assurance, open-source coding standards (Python 3, Docker, React JS) are used, data exchange standards are fol lowed when possible (XTCETM, C2MS), and compliance with international standards is followed (European Space Components Coordination, European Cooperation for Space Standardization, Consultative Committee for Space Data Systems, and others).

During the final stages of ESTCube-1, a difficult period marked the core design decisions for ESTCube-2. The launcher provider offered an earlier launching date. The ESTCube team decided to speed up the development time to make it into the new launching date —the CDS version of ESTCube-1 was selected to fit into Arianespace’s Vega rocket. In the end, everything went well except for one thing: the tether of the main payload failed to deploy on orbit, and the principal experiment did not take place.

As a consequence of the shortcomings detected in the ESTCube-1 subsystems organization and the critical failure on the principal payload, ESTCube-2 follows a unified organizational structure. It aims to maximize the use of common components, allowing reusability and “facilitat[ing] mobility of team members between subsystems” (Slavinskis, Pajusalu, et al., 2015). This integrated-system approach reduces the number of subsystems and improves the resilience and stout ness of the hardware. Building all subsystems as independent components connected only by cables resulted in an overall weak system structure in ESTCube-1. The ESTCube-2 orga nization of work is the same as ESTCube-1: the satellite plat form is developed in Tartu, while the payloads (one in ESTCube-1) are developed in specific places. The main difference is that ESTCube-2 is a 3U (three-unit) CubeSat and can carry more payloads. ESTCube-2 teams are based on dif ferent academic institutions (TO, UT Institute of Physics, Aalto University, Dresden University of Technology, Ventspils University of Applied Sciences and the Finnish Meteorological Institute). Each team focuses on a single pay load. Each payload is related to the expertise of the academic institution in which each team is based.

This technology-determined organization of work resulted in one more important defining trait of ESTCube: open and efficient personal communication skills. Different subsys tems demand expertise from different scientific and technical fields. Most ESTCube members tended to use jargon and technical terminology when communicating their work and findings, which complicated both the coordination between subsystem teams and the communication with external par ties. To solve this issue, ESTCube members take communication skills training and are generally encouraged to publicly talk and write about their work (Olesk, 2019). For internal communication, like task coordination or decision making, ESTCube first adopted an Estonian communication platform called Fleep (Liibert, 2017) and later moved into the open source oriented Discord platform. Using channel lists to organize working tasks and topics, every member can find the relevant information and documentation and take part in the conversations at any time in the same domain. "

(https://journals.sagepub.com/doi/abs/10.1177/02704676211041900)