The aim here is to identify areas where communication problems contributed to NASA’s inability to prevent the Columbia shuttle disaster, and to explore the benefits of the project management model outlined in the book and here for project management. The aim is not to assign responsibility or lay the blame on specific individuals. In fact corporate failure is never the fault of a particular individual; the whole organizational structure plays a role (as was noted also by the Columbia Accident Investigation Board). Large organizations are made up of multiple hierarchical levels, each consisting of teams who are specialist in their own fields of knowledge – organizational sub-cultures. These sub-cultures have their own values, priorities, styles of communication and discourse conventions. The goal of any study should not be to judge these diverse groups, but to find ways for them to communicate effectively with each other so as to contribute equally to achieving organizational objectives.
1. Make a narrative analysis of this story. If the participants in this situation had used this kind of analysis would it have clarified some issues, and made it easier to plan a course of action?
To form a narrative model, we will take the point of view of the Debris Assessment Team (DAT), since the story, as recounted here, unfolds around their actions more than the actions of any of the other participants. In our hypothetical scenario, DAT should have formulated a narrative analysis of the situation as soon as they were formed as a team. This would have enabled them to plan their course of action systematically.
The Event in this story is the debris strike that took place during launch. DAT should describe this in writing as fully as possible, with the available data. This would help them to clearly formulate the questions that require more data to be answered, and the main task that the team needs to perform – to get more imagery of the shuttle wing. A clear formulation of the questions would also make it easier for them to ‘translate’ these questions and other technical data for a wider audience.
Writing the Event in clear terms is crucial in identifying the steps needed to deal with it effectively. If the Event is not clearly conceptualized and outlined, suitable actions may not be easily identified, leading to another, possibly more serious, Event. In the Columbia case, not identifying the debris strike as an Event led to the disintegration of the shuttle, which became the Event for the Accident Investigation Board project.
The Controlling Scene
From the information that our story provided, we can trace the following Controlling
Debris Assessment Team
: a special team comprising engineers from NASA and the United Space Alliance, charged with analyzing and assessing the damage incurred by the foam strike.
Department of Defense:
contacted by NASA when military technology is needed. Protocol requires that this should be done through established procedures only when there is “mandatory need”.
Intercenter Photo Working Group
: analyzes video and photos from shuttle missions.
Johnson Space Center:
provides Mission Control for the Shuttle program.
Kennedy Space Center
: NASA’s launch facility.
Langley Research Center:
provides the Shuttle Program with structures and materials.
in large scale projects involving the future of humanity and the public interest (not to mention public money), the media is important, both as a source of support and as a constraint, and should be taken into account in project planning.
Mission Evaluation Room:
evaluates the flight from an engineering perspective and acts as intermediary between the engineers (and DAT) and the Mission Management Team.
Mission Management Team
: lead the project and are responsible for making the key decisions throughout the whole mission; they report directly to Shuttle Program managers. DAT members should involve the Mission Management Team in all major decisions.
all government agencies and corporations where public welfare and/or finances are involved should include the public in their planning. This could take the form of demographic or psychographic analyses, analysis of the composition of specific groups, etc, as appropriate (see Chapter One).
Space Shuttle Program:
responsible for NASA’s policies and procedures for space shuttle operations. If DAT are unable to obtain information from their direct leaders (Mission Management Team) on protocol, they should appeal to Space Shuttle Program members.
United Space Alliance
: a joint venture between Boeing and Lockheed Martin established to perform the Space Flight Operations Contract that conducts the day-to-day operation of the space shuttle.
The Controlling Scene includes a series of Roles, embodied by the following individuals (listed alphabetically by surname):
– head of Space Shuttle Systems Integration at Johnson Space Center.
– Johnson Space Center engineer, DAT member
The Crew (Rick Husband, William McCool, Michael Anderson, David Brown, Kalpana Chawla, Laurel Clark, Ilan Ramon):
it is important to list them as part of the controlling scene, because their individual expertise and skills could be useful at specific points in the Treatment. Notice how in the story (both as recounted here and in the actual chain of events), the crew were barely involved, even though their lives were at stake.
– Langley Research Center engineer.
– Shuttle Program manager at Kennedy Space Center.
– chair of the Mission Management Team.
– co-chair of the Debris Assessment Team.
– Mission Evaluation Room manager.
– Chair of NASA’s Intercenter Photo Working Group.
– co-chair of the Debris Assessment Team.
– Thermal Protection System specialist, Johnson Space Center.
– Manager at the United Space Alliance.
These roles should also include information on reporting procedure (who reports to whom), and individual responsibilities as they relate to the project. Since this is a government project, information on individuals’ levels of security clearance (and therefore of access to data) should also be included.
Besides groups and roles, the Controlling Scene should include all protocols or regulations that prescribe behaviour and that must be taken into account for processes to function smoothly. This should be done even if undertaking some of these procedures does not become necessary. In our story, we find this procedure:
- To obtain imagery of the shuttle in space, DAT must go through the Mission Evaluation Room to the Mission Management Team.
The Columbia disaster was in large part due to numerous unsuccessful attempts at finding out how to obtain imagery. Interestingly, DAT never went to the Mission Evaluation Room or the Mission Management Team for this, which suggests that they were unaware of the requirement. The Mission Management Team (if not, indeed, the Space Shuttle Program) should have made sure that DAT was given all necessary protocols; also DAT should have explicitly requested this from the Mission Evaluation Room, and then, if this was not successful, from the Chair of the Mission Management Team, Linda Ham.
Listing all relevant procedures deliberately and systematically helps the project team to anticipate and plan for eventualities, and also to identify missing but necessary information, and means of accessing it.
A final note
: DAT’s actual Controlling Scene would be much bigger (assuming they did their preliminary research thoroughly!) The Controlling Scene outlined above is based on the simplified story that we presented in this site.
The treatment of a project is decided after examining the nature of the Event and its requirements in relation to the constraints and resources of the Controlling Scene. Looking at what happened in the story, we can retrospectively trace these actions as DAT’s treatment:
1. Bob White contacted Lambert Austin (not an appropriate contact) to ask how to get imagery.
2. Austin called the Flight Support Office at the Department of Defense and asked for information on getting imagery (no follow-up seems to have been planned or done).
3. DAT held its first meeting.
4. Rodney Rocha contacted the Engineering section at Johnson Space Center to request imagery (not an appropriate contact).
5. Lambert Austin notified Linda Ham of the requests for imagery (leading to their cancellation).
6. DAT held second meeting to discuss the cancellation of their request (they did not invite Mission Management Team representatives).
7. Rodney Rocha sent a series of e-mails to engineers (not an appropriate contact) expressing concern (no plan was formulated on how to communicate urgency and priority to project managers).
8. DAT held a third meeting to review analyses.
9. DAT presented results to Mission Evaluation Room manager (the medium of information transmission – PowerPoint presentation - was inadequate, and no urgency concerning safety of flight issues was expressed).
10. Feeling dissatisfied with what had been happening, DAT contacted Bob Daugherty for simulations.
Clearly, these actions were inappropriate and unproductive. DAT seems to have tried to by-pass their project leaders, the Mission Evaluation Room and the Mission Management Team and tried to solve the problem within their in-group of engineers. Since a major component of the problem was administrative and not technical (permission and means to obtain imagery), finding a means to include the administrative managers right from the start should have been a priority.
The most appropriate course of action would have been for DAT Chairs to ask the Mission Management Team for clear protocol before doing anything else. They should also have contacted the Intercenter Photo Working Group and coordinated their efforts for imagery (notice that they did not even seem to know that the Intercenter Photo Working Group had set a process in motion with the Department of Defense). Also, they should have included representatives of the Mission Management Team in the second meeting and expressed clear concern to them about safety of flight issues. Finally, they should have persistently followed up every request they made and asked for specific feedback each time. This involves setting deadlines for obtaining feedback, and if no feedback comes by the deadline, there should be contingency plans to contact people further up the management ladder immediately.
2. Identify the areas where miscommunication occurred, and check them against the list of common failures in communication listed on pages 293-295 of the book. What went wrong? How could it have been avoided?
This project demonstrates several common failures in organizational communication. For example,
Ignoring the company’s role distribution.
Although obtaining imagery was vital for the success of the project, project team members did not know how to achieve this and did not persist in finding out. Also, they were not given any direction or guidance by their managers with regard to protocol and established procedures, nor were they asked for feedback on how they were progressing in their task.
The Mission Management Team was not involved enough in the Treatment, and when Linda Ham was eventually told of the request for imagery she could not find the responsible person or ascertain the reason. This is clearly a symptom of blocked channels of communication.
The requests by Wayne Hayle and Rodney Rocha for imagery were both made to inappropriate contacts. Also, DAT did not contact the Intercenter Photo Working Group for assistance, although they were the most likely source of support. Communications seem to have been one-way only with inadequate feedback and follow-up: for instance, the Department of Defense did not notify DAT or the Intercenter Photo Working Group that they had initiated action to obtain imagery, inducing the engineers to keep looking for alternate sources.
In all, there was a tendency for DAT engineers to communicate within their own group and leave out other groups that, in fact, would be more suitable as sources of support. In large, multi-levelled corporation, peer-oriented communication can often be subordinated by management decision-makers because they may interpret it as a non-critical, specialist task that involves issues solvable within the specialist team. At the same time, there was no move from the other groups to meet the engineers half way.
Not following the rule of ‘one document one message’ / Being unconvincing or irrelevant.
Another major source of miscommunication was inappropriate or vague wording of messages. In our story, when Lambert Austin contacted the Department of Defense, he did not emphasize the importance of the imagery for safety of flight. Instead, he said he was asking for information because some engineers requested it. This vagueness in the message invited it to be interpreted as an almost whimsical request for data to satisfy engineers’ curiosity, or perhaps for further research or for shuttle maintenance. Austin did not ‘own’ the request by personally endorsing it, nor did he state or imply that the data could involve matters of life and death.
When some engineers got an opportunity to voice concern after Linda Ham specifically asked them if there were flight of safety issues, they missed it by concentrating on technical analyses of available data and on speculation based on past experiences (which involved successes). A more appropriate approach would be to take a more pessimistic view (as is advised when safety matters are concerned), and explain the ‘worst case scenario’, what is likely to happen if there is serious damage to the wing. This was also an excellent opportunity to clearly articulate the need for imagery for reaching more accurate predictions.
A similar situation occurred when Bob Daugherty sent his simulations to selected groups. He sent his ‘worst case scenario’ to people who could do nothing about it (peers), and a sugar coated version to people who could take action. This is quite a common response by technical specialists, who do not want to appear too certain about situations where a degree of uncertainty exists. This is not a weakness: in fact, it is a commendable trait and part of scientific devotion to objectivity, thoroughness and accuracy. Exaggerations and scare tactics are not conducive to productive management operations and they lead to loss of credibility and a weakening of team spirit. However, when matters of safety are concerned, the best action is for all members of an organization to agree on how risk should be assessed and how this assessment should be communicated so that relevant variables are made clear to all parties concerned.
Another opportunity was missed when DAT presented their results to the Mission Evaluation Team. Relying on a PowerPoint presentation was a tactical error in communication because the information presented in this medium is much more likely to be interpreted as routine (given the number of PowerPoint presentations used in routine business meetings and briefing sessions in most corporations). As most of the important data was crammed into one slide, this was also a misplaced attempt to be concise at the expense of clarity. A report would have given the engineers more room to clarify and explain important issues and to emphasize points by positioning them strategically in the document.
An area where language played an important role in misunderstanding was in the interpretation of the policy clause stating that requests for Department of Defense participation had to be based on ‘mandatory need’. Since administrative managers often fall back on policy wording, it is advisable for the engineers to use this term for a situation where, in their own culture the wording may have been ‘obvious need’. By translating ideas into language that their audience can understand, writers/speakers maximize their chances of being heard, and of their message being received more favourably. This is a case of writer-reader complicity as discussed in Chapter One.
Another similar point, which we did not mention in the story but that is relevant to this discussion, is the naming of the team as the Debris Assessment Team. This name proved to be an unfortunate choice. Although DAT’s duties were similar to those of what would normally be called a Mission Evaluation Room Tiger Team, a NASA-established unit provided for in Shuttle Program protocol, the fact that DAT was not classified as a Tiger Team meant that the only formal procedure it was instructed to follow was keeping the deadline for submitting findings. This suggests that not naming the team a Tiger Team was a tactical error, depriving the team of an infrastructure. It also shows the importance of language, and the systems of classification that it grounds, in recognizing and ‘owning’ specialist units in organizational settings.
Finally, it is interesting that all the written communication that occurred in this story took the form of e-mail and PowerPoint. Both document types are characterized by brevity and screen-based transmission. In addition, both types are considered incomplete: e-mail is commonly seen as reflecting oral speech (note that most e-mail messages here involved a question or an answer to a question), and PowerPoint presentations are intended to accompany actual oral speech. All these characteristics encourage the writer to present information summarily, and constrain him/her from expanding or elaborating on a point, and from effectively combining language with visuals. Choosing the most appropriate document type for information is a strategic move and part of good audience dynamics.
3. Analyse the writing in the linked e-mail messages, using the concepts of ‘writer-reader complicity’ and ‘audience dynamics’ described in Chapter One, and the aesthetics of style described in Chapter Three. What was effective and/or ineffective about the style and organization of these messages?
E-mail A -extract (sent by Rodney Rocha to Johnson Space Center engineers, on 21 January 2003, to request imaging by external sources)
The meeting participants […] all agreed we will always have big uncertainties in any transport/trajectory analyses and applicability/extrapolation of the old Arc-Jet test data until we get definitive, better, clearer photos of the wing and body underside. Without better images it will be very difficult to even bound the problem and initialize thermal, trajectory, and structural analyses. Their answers may have a wide spread ranging from acceptable to non-acceptable to horrible, and no way to reduce uncertainty. […]
This is a relatively effective message to peers. The first sentence states the issue clearly and unequivocally. The second sentence describes the effects on the problem caused by a lack of data. The third sentence is the most ambiguous, mostly because of the use of the word ‘answers’, but we can reasonably assume that a peer audience would make appropriate contextual inferences.
A strength of this e-mail is the use of emphatic language: ‘all agreed’, ‘always’, ‘big uncertainties’, ‘definitive, better, clearer’, ‘horrible’. Although such wording is generally inappropriate for formal writing, it is effective here since it creates a tone of urgency, which is exactly what was needed to convey the message. The major problem with this e-mail is that it ‘preaches to the choir’, describing some issues to an audience that would not contest them, but at the same time would not have the power to do anything about them. In fact, we could say that this e-mail creates writer-reader complicity with the wrong readers. It is precisely the tone of urgency found in this e-mail that should have been communicated to administrative managers.
E-mail B (sent by Calvin Schomburg to Ralph Roe, on 22 January, in response to Linda Ham’s question if there was a flight safety issue)
No-the amount of damage ET foam can cause to the TPS material-tiles is based on the amount of impact energy-the size of the piece and its velocity (from just after pad clear until about 120 seconds –after that it will not hit or it will not enough energy to cause any damage)-it is a pure kinetic problem – there is a size that can cause enough damage to a tile that enough of the material is lost that we could burn a hole through the skin and have a bad day – (loss of vehicle and crew – about 200-400 tile locations (out of the 23,000 on the lower surface) – the foam usually fails in small popcorn pieces – that is why it is vented-to make small hits- the two or three times we have been hit with a piece as large as the one this flight- we got a gouge about 8-10 inches long about 2 inches wide and ¾ to an 1 inch deep across two or three tiles. That is what I expect this time-nothing worst. If that is all we get we have no problem-will have to replace a couple of tiles but nothing else.
[ET= External Tank, TPS= Thermal Protection System]
This is an ineffective and poorly written e-mail. The first sentence consists of 174 words and contains enough information to write at least one paragraph of 5-6 sentences. The use of dashes and parentheses is inappropriate, not only because it expands the sentence, but also because it de-prioritizes information, producing a disorganized flow and stream of consciousness effect with inadequate pauses or sequencing. Parentheses especially are confusing because they subordinate information, while here we have some information that clearly is not meant to be subordinated. The phrase ‘loss of vehicle and crew’, for example, is put in parentheses and between dashes, so it is impossible to tell what verb it goes with, or if it is the subject or object of a clause.
Such a confusing message with no signposting or highlighting produces bad audience dynamics. The reader tries to extract the main point, and is likely to take this to be the word ‘no’, which can be inferred to be the answer to Linda Ham’s question if there were safety of flight issues. The rest of the message contains a technical description of what might go wrong, but does not specify that this description is constructed from speculations made on available data and past experience. In other words, it is once again an optimistic interpretation of risk. The last sentence is a conditional clause, ‘if that is all we get we have no problem’, showing uncertainty; however, since this comes at the end of a set of statements that do not contain any caution, its impact is greatly reduced and it is ineffective as a warning.
The e-mail clearly has not been revised: information is not chunked according to importance and there are several grammatical and stylistic errors. For example, a verb is missing in the information segment after the first dash, ‘will not enough energy’, and ‘that’ is repeated three times in the information segment after the second dash, ‘that can cause…that enough…that could burn’.
E-mail C (written but not sent by Rodney Rocha after imagery requests were cancelled)
In my humble technical opinion, this is the wrong (and bordering on irresponsible) answer from the SSP and Orbiter not to request additional imaging help from any outside source. I must emphasize (again) that severe enough damage (3 or 4 multiple tiles knocked out down to the densification layer) combined with the heating and resulting damage to the underlying structure at the most critical location (viz. MLG door/wheels/tires/hydraulics or the X1191 spar cap) could present potentially grave hazards. The engineering team will admit it might not achieve definitive high confidence answers without additional images, but, without action to request help to clarify the damage visually, we will guarantee it will not. Can we talk to Frank Benz before Friday’s MMT? Remember the NASA safety posters everywhere around stating, ‘If it’s not safe, say so’? Yes, it’s that serious.
[SSP=Space Shuttle Program, MLG= Main Landing Gear, MMT=Mission Management Team]
This is arguably the best piece of communication in the story, and it is disappointing that it was never sent to any responsible parties. This e-mail states clearly and emphatically the nature of the problem (second sentence), and connects it to the importance of obtaining imagery in order to better assess the problem’s seriousness (third sentence). Also, it begins with a clear evaluation of the decision not to obtain imagery (‘wrong and bordering on irresponsible’), and ends with a statement that reiterates the safety nature of the problem and its gravity (the last two sentences). The reference to NASA ideology (‘remember the NASA safety posters everywhere around stating, ‘If it’s not safe, say so’?’) is strategically appropriate because it makes a connection between the e-mail’s main point (‘yes, it’s that serious’) and existing knowledge, therefore showing relevance. In contrast to e-mail B, this e-mail rightly stresses the potential negative outcomes of the debris strike.
The obvious problem with this e-mail is that it was never sent, let alone sent to the appropriate parties. Also, the writing could be improved in some areas. For example, the self-deprecation is misplaced and does not add value to the message (‘in my humble technical opinion’). In addition, the third sentence expresses unnecessary hesitation in its negative structure (‘the engineering team will admit it might not achieve definitive high confidence answers without additional images, but, without action to request help to clarify the damage visually, we will guarantee it will not’). This sentence begins with a concession aimed at an anticipated criticism (‘the team will admit’) that mitigates the strength of the proposition; also, the phrase following ‘but’ does not follow the topic-action rule described in Chapter Three, and is convoluted and non-specific (‘without action to request help to clarify…’). An improved revision would be: ‘The engineering team needs additional images to achieve definitive high confidence answers. Without these it cannot ascertain the damage’.
Of course, any attempt at revising this e-mail should take into consideration the audience and their requirements and expectations.
This e-mail is emphatic and expresses the seriousness of the problem in its tone of urgency. However, like all the e-mails considered here (and, indeed, the pieces of communication used in this story in general), it does not recommend specific actions, and is, therefore, more descriptive and evaluative than directive.
E-mail D (sent by Bob Daugherty to Carlisle Campbell, in response to the footage he sent showing the debris strike)
I bet there are a few pucker strings pulled tight around there!
Thinking about a belly landing versus bailout…(I would say that if there is a question about main gear well burn thru that its crazy to even hit the deploy gear button…the reason being that you might have failed the wheels since they are aluminum…they will fail before the tire heating/pressure makes them fail…and you will send debris all over the wheel well making it a possibility that the gear would not even deploy due to ancillary damage..300 feet is the wrong altitude to find out you have one gear down and the other not down…you’re dead in that case)
Think about the pitch-down moment for a belly landing when hitting not the main gear but the trailing edge of the wing or body flap when landing gear up…even if you come in fast and at slightly less pitch altitude…the nose slapdown with that pitching moment arm seems to me to be pretty scary…so much so that I would bail out before I would let a loved one land like that.
My two cents.
Since this is peer-directed communication, the style and technical content of this e-mail as well as its informality are acceptable. The e-mail addresses directly the concerns raised by the footage and would therefore provide some useful information to the recipient, who initiated the communication.
However, the writing poses some problems. First, the humorous tone is inappropriate in the context of the gravity of the situation that the e-mail describes, and gives it almost an ironic hue (it is not clear if the writer is worried or amused). Second, like e-mail B, this e-mail is not revised and the content is not structured into correct sentences; the writer uses ellipsis marks to separate one segment of information from another and encompasses about half the message between parentheses. Therefore there is no emphasis, highlighting or sequencing.
Also, although the e-mail includes both description and evaluation, it mixes the two, making it conversational and therefore relatively insubstantial (for example, the clause ‘you’re dead in that case’ evaluates the technical description that preceded it). Most seriously, although the e-mail ends with a very worrying remark, ‘I would bail out before I would let a loved one land like that’, no specific action is recommended.
A more effective strategy to present this information would be to begin with a technical analysis of the footage, then to proceed to an evaluation of the seriousness of the situation, and end with a set of specific actions that would help to minimize the risks.
© Macmillan Publishers Ltd. - Houndmills, Basingstoke, Hampshire, RG21 6XS, England