What can we provide for Integration Engineers to improve their decision-making processes during launch countdown?
Skills: Interviews, Sequence Modeling, Affinity Diagramming, Visioning, Storyboarding, Speed-Dating, Paper Prototyping, Wireframing, Development, Think-Alouds
Duration: 7 months, 2016-2017
Link: Project Site
We dove into research with a wealth of publicly available material regarding NASA, literature pertain- ing to our analogous domains, competitive analysis, and academic research journals/books in the fields of cognitive science and psychology. We interviewed professionals from analogous domains and a former employee from SpaceX. We also conducted phone and in-person interviews with a wide variety of NASA stakeholders, from big picture-oriented Knowledge Management officials, to operations-focused engineers, to an astronaut. After narrowing our scope and identifying one particular scenario of interest, we travelled to Kennedy Space Center, where we interviewed the primary stakeholders at the launch center while immersing ourselves within the context (left).
As it was impossible for us to interview Integration Engineers during a real launch countdown, we turned to analogous domains for aspects of our research. These domains also broadened our perspectives and enabled us to benchmark and make comparisons.
Live TV Production, as it requires the coordination and collaboration of many specialists who work under a sensitive timeline towards the singular goal of airing the show live. This includes a complex web of actors, including a television director who oversees the audio, graphics, camera, and replay teams, and who also manages the expectations of the producer and announcers.
Emergency Medical Professionals make critical life or death decisions under significant time pressures and cognitive constraints. Understanding their cognitive flow within information-dense, high-stress environments provided insights on the role of human error in misdiagnoses, and furthermore, the impact of cognitive load and biases leading to critical mistakes.
Lastly, we looked into Business Information Modeling (BIM), a tool used by the Architecture, Construction, and Engineering industry. BIM enables a 3D digital information artifact to be handed from team to team, while increasing collaboration and reducing knowledge loss.
In order to deeply understand the Integration Engineers' needs, it was critical that we intricately understood their roles in the launch room and the roles of the people with whom they collaborated and on whom depended. Therefore, we interviewed people from each of these subgroups. During one interview, we sketched a simple map of the launch room (below). As you can see in the sketch, the Integration Engineers are seated at the center of the launch room, surrounded by the diverse System Product Teams that report to them. Later, we created a Stakeholder Map (left), depecting the relationships in more detail.
The 5 of collaborated to combine and make sense of our extensive interpreted interviews, on-site sketches, reading notes, etc. We created visual representations of our shared findings to give us a unified jumping off point. This image shows our team during one of those sessions generating an Affinity Diagram. Ultimately, we distilled our research into 4 clear insights that informed our decisions throughout the design and development phases. These insights are described in depth in a book that we presented to NASA after the first 4 months of this project. It is available for your viewing upon request.
When we arrived at NASA after months of research, we presented our insights to the Human Factors team. Directly following, we led a visioning session, where we generated hundreds of possible solutions. Using those concepts as a starting off point, we continued ideating. We gave these ideas form through rapidly producing early prototypes. We evaluated the feasibility and impact of our designs and moved forward with the top contenders.
Over the course of the summer, we conducted 3 formal rounds of user tests. Sometimes we also led more casual tests when groups of people would stop by to see our prototypes that they had heard about. We made a scenario for each test and paired it with live audio or recordings, to simulate the operating system in the launch room. These tests helped us refine our prototype in future iterations.
Our final prototype provided a solution for each of 3 problems that we identified during our research, as depicted in this timeline of an anomaly call in the launch room during countdown. Visit the results section to read more about each solution in greater detail!
In Kennedy Space Center’s firing room during launch countdown, unexpected problems occur. Integration Engineers are responsible for troubleshooting these anomalies using audio and visual information. Anomalies are primarily communicated over an intercom system. The Integration Engineer must simultaneously access NASA’s vasts webs of data to ascertain the problem and collaborate to determine a course of action. To complement and improve this process, our prototype, Loop, bridges the gap between audio and visual information.
The primary goals of Loop are threefold: (1) alert the Integration Engineer to important announcements or events, (2) reduce cognitive limitations inherent to listening to several audio channels at once, and (3) present the most relevant data.
Currently, critical events are only announced over the intercom system, where they must compete for the attention of the Integration Engineers with the cacophony of many less significant conversations happening simultaneously. Conversely, Loop uses prioritized notifications to provide a redundant and prominent avenue to receive this essential information. Notifications grab the Integration Engineers' attention to present information they need, right when they need it. They also encourage Integration Engineers to take action by providing the context they need to move forward and buttons on the notifications serve as entry-points into the data.
Loop only aims to pull the Integration Engineers' attention away from their complex and vital work for 5 categories of critical events that they can not afford to miss. “Break-break” alerts are of the highest importance and command the launch room team to stop whatever they are engaged in, as there is a critical issue that must be resolved. Anomaly calls, the second-most important type of notification, denote the start of the troubleshooting process led by Integration Engineers. Lastly, events such as call-sign (each person's title in the launch room) mentions, milestones (e.g. start of fueling), and weather conditions (e.g. lightning or dangerous winds over a certain threshold), also demand immediate attention.
To prevent Integration Engineers from missing valuable information as they often do, Loop transcribes all communication over every intercom channel. The various transcriptions, serving as a point of reference whenever necessary, are easily accessed and navigated. The text includes buttons that direct Integration Engineers to relevant data when clicked. Details in the scroll bar empower Integration Engineers to scan for specific details that are related to their current tasks. Additionally, NASA personnel find customization techniques to be reassuring and helpful during the stressful countdown until launch, so Loop enables customization with keywords. Integration Engineers can preset significant words and phrases to highlight in the transcription, rendering them easily searchable and identifiable.
NASA generates an exorbitant amount of data on every aspect of the vehicle. Integration Engineers need to access this information when troubleshooting an anomaly. Loop’s data integration in the Dashboard empowers Integration Engineers to interact with disparate data presented succinctly and intuitively, uncovering relationships between related technical requirements and real-time telemetry sensor data from the spacecraft and rocket. Within this unified, customizable environment, Integration Engineers can move onto in-depth analysis, status discovery, and towards finalizing a decision to launch.
Loop’s Dashboard serves Integration Engineers with the most important information, but only in short summary; it’s the tip of the iceberg in the vast amount of data Integration Engineers may eventually access. Rather than overwhelming them with all of it initially, each element on the page is exploratory in that it can be clicked on to retrieve further explanation. It is up to the Integration Engineers to maneuver to which aspect they deem most important. Additionally, hierarchy is foundational in the design. The organizational structure of the presented data aligns with the data’s actual structure, matching the Integration Engineers' mental models.