Chapter 6 - Results

In this first batch of results, we task ourselves to demonstrate an anomaly in the presented case studies by opposition to the literature models in our research questions chapter. The results are presented in the same fashion for the three selected case studies. They are also detailed in different articles in the publications.

Systematically, we present the trajectory of the project where we stress what appears to be a weak signal with respect to generative processes at stake. It appears as an obstacle for collective action, something that has created a management issue for involved designers, engineers and managers. We then propose to use the adaptive, interactive and encapsulated models to predict what could have been the course of action. It will serve as a reference point for the actual course of action and detect the anomaly with more precision as explained in the methodology section.

---

Icing conditions: misunderstood entrepreneur or absurdity?

Trajectory of the project

As described previously in the methodology sections, the case was suggested to the researcher for analysis as the R&T manager had managed the project from the beginning and had increasing pressure on the conditions of exploitation after the years of exploration and financial constraints to continue exploration.

The figure.1 below, gives the chronology of the project as it was developed over more than two decades. The sequence can be followed through the numbers at the corner of several boxes and years highlighted at the bottom. The arrows indicate sequence and causality operating for actions, decisions and the design of decisions. It is, of course, a condensed view of the history of the project but we can stress the nature of exploration (disrupt monopoly), when exploitation (client request and market opportunity) could have occurred and when exploration was launched again.

The vertical dotted line is the critical part in the project trajectory. A first exploration was carried out by the project team looking for alternative solution overcoming limitations known of the monopoly’s solution. They scouted the field, several test campaigns were conducted with the support of national laboratories and aircraft manufacturers. These were proven successful despite some caveats as the first technology developed addressed only ice detection on ground. However, another solution was developed with further test campaigns. Regional aircraft manufacturers were looking for an alternative solution to the monopoly and possibly for an integrated system: detection and protection from ice. Restrictions were looming over regional aircrafts as they were statistically known more sensitive to icing conditions.

Surprisingly, instead of going for this aircraft program and system/product development between ZFCS and ZEEU that was at hand’s reach, the current technology development were stopped altogether. Exploitation promises were discarded, in order to launch a new wave of exploration addressing yet to be specified icing conditions: new undefined regulations, scientific knowledge on icing conditions, and appropriate technology. We propose then to zoom on this key moment of the project history that balances exploration and exploitation in order to make sense of it through the lens of literature models.

Models’ predictions

Adaptive model predictions for Icing condtions

Model descriptors	Predictions of the adaptive model
Model of coordination and collective action	BU management and group top management were involved in key decisions (search direction for technology development, patenting, business acquisition, etc.). Prediction: BU management could have selected the program opportunity with the development of the system between ZFCS and ZEEU as a means to set a first foot on the monopoly for future learning promises in the new environment.
Generative processes	The exploration was rich in finding existing alternatives but generating new ones that are undecidable is not considered by adaptive models. Prediction: The exploration movement should have been gradually adapted to fit exploitation constraints. Make or Buy strategy for the detection technology (e.g. acquire Vibrometer firm) in order to bring value added with an integrated system (detection and protection) by opposition to a standalone solution.
Environment cognition	Interactions were created with laboratories, test facilities and potential clients to guide the search direction: overcome limitations of monopoly’s solution. Prediction: Wait for the environment to stabilize and prescribe new requirements on icing conditions regulations, which can make the generated concepts decidable.
Organization design	R&T team dedicated to exploration, spatially isolated from engineering department. It was identified as a new standalone product line. The development engineering was familiar with systems engineering. Prediction: The exploitation could have been conducted for the icing conditions detection and protection system between ZFCS and ZEEU. Resources were available.

Interactive model predictions for Icing conditions

Model descriptors	Predictions of the adaptive model
Model of coordination and collective action	R&T engineers and managers created and made sense of the surveyed technologies. Meetings were organized with middle and top management to structure future course of action. Prediction: Middle and top management allow and enable further interactions to gain awareness and act accordingly.
Generative processes	Test campaigns, studies and meetings were conducted and shared among engineers and managers. Presentations are held with management to envision new positioning Prediction: The appropriate sense and knowledge could have eventually captured through subsequent interactions. R&T manager could have then refined the new meaning emerging from further exploration.
Environment cognition	Numerous interactions with clients, laboratories and potential clients were conducted to address uncertainties of icing conditions and required technologies. Prediction: Shape new networks, lead-users and unknown interpreters to gradually design new technologies for icing conditions. Continue developing open innovation practice.
Organization design	The exploration team had expanded beyond the ZFCS boundaries and gradually gaining recognition. Prediction: This open innovation should have easily shaped a fully autonomous product line as the legitimation process is carried out. Perhaps even a new BU (ad-hoc) could be created.

Encapsulated model predictions for Icing conditions

Model descriptors	Predictions of the adaptive model
Model of coordination and collective action	The exploration was seen as sequenced projects probing alternatives’ validity and success. The overall icing condition detection theme could be considered as a program but without the management structure, despite having a governance with middle and top management sessions. Prediction: Road-mapping activities to pave the way for future projects.
Generative processes	Each specific project encapsulated design and engineering practices to refine the selected alternatives. Prediction: These trial-and-error projects proving the technology, regulation and market landscape should continue until an appropriate fit is crystallized.
Environment cognition	Each project engaged with clients and test facilities, and they allowed building upon each other. Prediction: Each project will explore the environment in a given direction and contribute to future projects. For instance, the program could have enforced the exploitation opportunity to sustain further exploration on the side.
Organization design	The projects relied on R&T engineers and request additional resources if needed. Prediction: Transferring the project to the development engineering to develop the integrated system would have been possible due to system related knowledge.

Anomaly: market legitimacy for the market but no organizational support

Model descriptors	Actual course of action
Model of coordination and collective action	The team conducting exploration didn’t support the exploitation opportunity but rather continued exploration. They handed over some R&T engineers to develop “on the shelf” the first version of the explored product, whilst further exploration continued shaping the unknown (norms and regulations, and aircraft integration and safety)
Generative processes	The exploration was guided by a future prospect of exploitation: safety norm evolution and associated icing conditions typologies. They willingly reversed preferences by choosing a highly uncertain technology, but managing the environment legitimacy in parallel.
Environment cognition	The environment is shaped by the R&T manager constituting a collective path for the market: deciding regulation’s scope and application, steering scientific discoveries and applied research on icing conditions, and specifying norm and technological performance.
Organization design	The projects managed by the R&T team were rather autonomous with partnerships and funding agreements. They gained autonomy up to a point where developing a product required to hand over some R&T engineers to the Product Development department.

What our descriptors reveal is that the conducted project does partially fit some of the models with an emphasis on the interactive and encapsulated models. The team engaged with the environment’s actors and pushed the trial-and-error pattern to ensure the legitimation of the developed technology as well as the primacy/leadership of the newly reconfigured icing conditions detection market.

However, we see that the exploration is guided by exploitation constraints that still requires to be fully settled and designed. The project is at odds with organization design as they have evolved in autonomy and developed capabilities so far unmatched by legacy engineering departments. Top management, despite having sponsored the exploration, still tries to keep the BU together and avoids having a spin-off. As the technology is not fully mandatory for existing heavy-body aircraft (statistically proven by manufacturers), it still requires to make a case for itself beyond small/regional aircraft market. The BU strategy is thus challenged by this market re-orientation and the lack of technology-market fitness.

Lone irrational entrepreneur?

One could argue the R&T manager used foolishness (March 2006) to refuse the first exploitation opportunity benefiting from synergies with another BU. He is however a sane spirit in a sane body.

The adaptation of technology is performed dynamically with the environment being shaped by managing safety regulation and norms as well as technology legitimation. Exploration of technologies and product concept are guided by generating and refining alternatives of what exploitation (regulation and standard) could be.

The non-mutual conditioning of the regimes is compromised, but brings new hopes for exploitation. As discussed in (Le Glatin, Le Masson, and Weil 2017 b, see publication), preferences were reversed, representing a violation of rational theory of choice, but at the same time allowed uncovering prospective alternatives requiring a contingent environment management. This entrepreneurial effort is not isolated, it was sustained through a rich engagement with the environment, learning for the R&T team, but also support from BU management and top management.

New BU for ZSSM?

This offshoot generated by the R&T team dedicated to icing conditions detection was established as a young product line. The issue for top management is the absence of a clearly defined market for heavy-body aircraft, which are the norm for other products in the BU.

It is probably the paradox of this exploration project and the autonomy gained through open innovation practice, path constitution and funding schemes. It is abnormal for the three models, as the top management sponsored the emergent strategy from the beginning, and steered the process along the way as a strong technological advantage could be developed to overcome the monopoly.

In the end, exploitation is shifted from its standard belief and conception to market the product, from a marketing and engineering standpoint. In fact, the prototype developed and tested by R&T implies a lot of rework for development engineering, in terms of robustness, quality and manufacturing engineering. Nevertheless, the prototype can also be seen as a means to support learning of the new technology so far unfamiliar to the engineering department.

Creating a new BU appeared out of the equation. BU management preferred to capitalize on the exploration and develop the new technology on the shelf. They also allowed with external funding to let the new technology version to be further explored as the norms and regulations being co-managed by the R&T manager have created new slots requiring applied research and product development to ensure aircraft safety and airworthiness.

---

A seat platform: poor exploration management?

Trajectory of the project

As described previously in the case presentation, the researcher was encouraged to have a closer look to this seat design after we had organized several meetings and workshops to support the BU in implementing their own management tools and practice to evaluate TRL for R&T projects and support handovers with Development Engineering. This Business Class seat platform project, not only raised the assessment issue for isolated technologies and its integration, but rather emphasised the criticality of design robustness as a system to meet requirements.

The figure 2 below, gives the chronology of the project as it was developed over more than two years. The sequence can be followed through the numbers at the corner of several boxes and years highlighted at the bottom. The arrows indicate sequence and causality operating for actions, decisions and the design of decisions. It is a condensed view of the history of the project but we can stress the nature of exploration (platform envelop definition and selection) and the shift towards exploitation requested by a demanding client.

The vertical dotted line is the critical part in the project trajectory. A first exploration was carried out by the project team looking for alternative solution overcoming limitations of packaging and certification for business class (BC) seat. Several options were found in the market (patents, competition benchmark) and they designed several alternatives before elaborating criteria with BU stakeholders to decide on the concept. They ended up with a seat architecture sharing a frame with neighbouring seat (like a bench) supporting their own load and a shell providing volume and several valuable BC features. The modular design puts the emphasis on the seat lower frame as being the core on which all other modules would plug in (foot rest, backrest, shell, etc.).

This is when sales and program management teams saw the attractivity of the product with a greater living space which seduced a demanding client. The concept design was then selected by the client and further refined with their own design studio. However, this shift from exploration to exploitation revealed several complications in the design architecture. For instance, the 3 legged seat and the load of both seat and shell packaged together as whole for certification (i.e. airworthy and safe for passenger) raised the mechanical engineering complexity. This a priori successful balance from exploration to exploitation came out to be quite disastrous: quality issues, delays, engineering cost overruns, and jeopardized seat platform.

Models’ predictions

Adaptive model predictions for BC seat platform

Model descriptors	Predictions of the adaptive model
Model of coordination and collective action	Sales, Engineering and Strategy managers took part in balancing out exploration and exploitation effort. Prediction: Management would actually encourage exploitation of the explored alternatives and subsequent decision-making (concept selection process )
Generative processes	The exploration was rich in finding and generating alternatives, ensuring enough room for engineering to develop a solution. Prediction: Once the concept is selected, the uncertainty reduction mindset of development projects would be applied to meet client requirements and product early-specifications (modularity).
Environment cognition	The technology and patent scouting was rich and allowed finding white spaces to develop new solutions. Known issues for packaging, aircraft integration and certification-ease were used to design alternatives. Prediction: Once a client can make a demand fitting existing offer (preliminary design), the exploitation should be launched and further development would be based on adapting the course of action.
Organization design	R&T team dedicated to exploration and isolated from engineering department. They were tasked to create a new product concept based on modularity and new requirements. Prediction: When exploitation requirements are given, the project should transfer its resources to development engineering.

Interactive model predictions for BC seat platform

Model descriptors	Predictions of the adaptive model
Model of coordination and collective action	R&T engineers and manager designed digital mock-ups, coordinated with stakeholders to select concepts. Prediction: Middle and top management pushes the mock-up and prototypes among department to learn by doing and meet requirements along the development process.
Generative processes	Tests, simulations and meetings were held around digital mock-ups to refine and generate choices. Prediction: The mutual interactions between individuals and prototypes during presentations, and client meetings will gradually ensure a smooth uncertainty reduction.
Environment cognition	Demonstrated awareness and scouting of the environment through presentations and consultations with BU stakeholders (engineering, marketing, packaging, certification). Prediction: Added requirements will be identified through interactions with client and prescribers (design studio). In exchange, the proposed seat platform features will be discussed and mutually adjusted.
Organization design	The initial R&T project spreads beyond the R&T team boundaries with involvement and consultation of different stakeholders. Prediction: Depending on the evolution and definition of requirements, the necessary resources will be found and support the product design. The Engineering department may have to be reconfigured to fit the modular design.

Encapsulated model predictions for BC seat platform

Model descriptors	Predictions of the adaptive model
Model of coordination and collective action	The exploration project had a concern for platform engineering feasibility and an exploitation target (packaging, ease certification). Middle management as well as BU management supported the initiative. Prediction: The project would be part of a product line or stream of projects, and eventually a program for business class seat platform.
Generative processes	The design practice to define the seat platform made an effort to define the minimum and common features required for modularity. The selection criteria also shaped the design of alternatives. Prediction: As new requirements would be demanded by the client, the project would gradually adapt and propose variations to the platform.
Environment cognition	The project reached out for unmet needs from value chain and aircraft integration to extend the project’s requirements. Prediction: The client’s project would define new requirements and induce constraints that would have to be adjusted to the R&T project’s specifications.
Organization design	The projects relied on R&T engineers and request additional resources if needed. Prediction: Transferring the project to the development engineering to develop the seat platform would require to temporarily find resources in engineering departments to solve conflicting requirements.

Anomaly: a rich exploration jeopardized by a demanding client

Model descriptors	Actual course of action
Model of coordination and collective action	The exploration is guided by exploitation objectives. The shift is capitalized by the client and its requirement. During the actual product development, mechanical engineering issues were faced due to novelty/unfamiliarity of the product for development engineers.
Generative processes	Careful exploration was conducted with criteria for selecting them defined with BU stakeholders. The complications faced by exploitation activities raise numerous questions on the evaluation of generated/selected alternatives. In the end, the platform design is no longer modular and the client’s product becomes specific and has generated cost overruns.
Environment cognition	After exploring the environment to define the platform, the client and its design studio imposed requirements forcing the platform design to be adapted. The built-up awareness for exploration is reduced and non-negotiated with the client (imposed requirements).
Organization design	The project gradually moved from R&T to Engineering, but also had to come back to some R&T engineers to solve complications in mechanical engineering practice. The Engineering Department struggled to meet requirements.

This comparison shows that the actual course of action is far more complex and raises again numerous obstacles the literature models would not have fully grasped. For instance, the decision-making process organized during the exploration phase was used not only to select a platform design, but also as a basis to generate new alternatives in between decision meetings.

The transition from exploration to exploitation was seen from the angle of the client who identified a feature (increased living space and leg space) in the platform seat (with its enhanced packaging and certification). However, the difficulties were treated by the Engineering department requiring to call back R&T engineer to support them. There is then an overlap of the influences between exploration and exploitation.

Lack robust design practice between R&T and Development?

One could argue that the engineering is overall of poor quality. It is an argument hard to sustain as the BU is one of the market leaders and pioneered several designs. Moreover, the product development is conducted in short time period (2 years approximately) and it requires a lot of flexibility within engineering departments to support the design and development. The researcher had the occasion on multiple occasions to witness the density of relationships and knowledge sharing within a project’s scope.

Together with the Engineering Director, ex-post reviews were conducted to evaluate the maturity and robustness of the design during the exploitation phase (client’s program gates). It revealed that the several mechanical issues relating to the interference of the platform design and client requirements were understood late in the development process. The researcher found that traditional product development (i.e. no modularity) relies a lot on the physical testing conducted quite late in the process due to parts availability¹. However, simulation test engineers had flagged several flaws of the design, but were not heard by development engineering preferring to witness the physical test, whereas they were by R&T engineers during the exploration phase.

Finally, it is not a complete failure as teachings have been transferred to R&T team and advanced concept team (group of internal designers and engineers) to study again the possibility to develop a new modular design, with more emphasis on the shell design. Another topic on modularity was also launched by the R&T team approaching the concept differently, for instance, not in seat-centric way.

Too early exploration/exploitation transition?

One could also argue the project was handed over to quickly to the engineering department. However, it is not really the case as the selection of concepts was made collectively and interactively with BU stakeholder. The client’s request for proposal and bid award was an actual great opportunity to value the exploration phase conducted by the R&T team.

What is perhaps missing is the weakness in countering the demanding client’s request which have pushed the physical limits of the platform design. Why not using the client’s opportunity and constraints to develop in parallel a full platform design instead of trying to keep some complicated associated features (e.g. totally new compact seat kinematic)?

The exploration and exploitation were put on a same continuum, being perhaps overlapped at the beginning of the project. Later, it unfortunately over-complexified the product design and development, to a point where the innovation effort in the platform design is killed by switching to a normalized exploitation regime. When discussing with engineers and managers why some alternatives had been discarded (designed by the researcher with C-K Theory, see (Le Glatin, Le Masson, and Weil 2018, see publication), signs of regret were identified explaining the difficulty to identify and articulate the exploration/exploitation divide earlier in design practices.

---

ADT/Design Thinking: resistance, death valley and organizational misfit by design?

Trajectory of the project

The last case we are presenting here is the synthesis of two projects managed by the Airbus Development Team. The researcher had the occasion to analyse two projects led by two different members (engineer and designer) who had been trained to Design Thinking. The other ADT participants had been also introduced, if not trained to it, like it was the case of ADT’s manager. The latter had given access to all documentation available on their server and requested his colleagues to save some time for my interviews. The team had been tasked to generate proposals valuing BU’s know-how and combined synergies in the form of Multi-BU projects. They usually referred to the recurrent argument of the not-invented here syndrome or resistance to change to explain the absence of follow-up by BUs to develop their generated concepts.

However, they stressed some cases where BUs were thrilled by having opened their eyes to the competencies of sister BUs in the group, unlocking interdependencies between products that are usually hidden by market segmentation and sales channels. The projects based on Design Thinking adapted the method to have BUs not only learn from user knowledge, but also to have multi-BU design workshops.

As presented in the publications on Design Thinking, the cases were analysed with the help of C-K design theory in order to understand the performance of the generative processes in the projects and their impact for BUs (ADT’s mission). The C-K mappings were synthesized from control groups of students from Mines ParisTech. The figure 3 below shows the synthesis of the projects’ trajectory. Both design briefs (Better Waste Management and Turn Around Time Optimization) followed the same pattern as they had defined a methodology canvas adapted from their practice, experience and Design Thinking.

Our focus is on the output of the projects as they come to a standstill. If Design Thinking may not be the key to everything, it is worth understanding the relationship between the nature of generative processes and their organization impact: concept appropriation and development by business units.

In our publications on ambidexterity, this transition from exploration to an absence of exploitation is analysed. Below, we propose rephrasing it with descriptors and models’ predictions derived from our literature review. The trajectory is summarized in figure 3, and they strech over less than a year (approximately 8 months).

Models’ predictions

Adaptive model predictions for ADT/Design Thinking projects

Model descriptors	Predictions of the adaptive model
Model of coordination and collective action	ADT manager leaded the balancing of exploration and exploitation prospects (BU’s interests). Sponsorship from the VP of Business Development & Strategy. Workshops were organized to balance out concepts and their rework. Prediction: Leadership and top management identifies how the concepts can fit BUs’ strategies.
Generative processes	The exploration was supported by Design Thinking generative power driven by user empathy and extreme user cases. Prediction: As user/client value is supported by Design Thinking practice, the product development (exploitation) should naturally occur.
Environment cognition	The environment and value chain was scouted with the help of Design Thinking and other ADT’s activities such as previous projects and technical support. Prediction: The newly explored environment should ensure enough decisional/problem-solving background for BUs to engage in exploitation of concept proposals.
Organization design	ADT team is an example of contextual ambidexterity organized at the group level. The exploration is delocalized, interactions with relevant BUs are maintained along the exploration process (for organizational learning). Prediction: Exploitation of proposed concepts will be slightly adapted to the products design as they may require contractual development between BUs. Organization design should mirror the product as they have accepted and participated to the concepts’ user/client value.

Interactive model predictions for ADT/Design Thinking projects

Model descriptors	Predictions of the adaptive model
Model of coordination and collective action	ADT members and BUs correspondents, at stake with project’s topic, are in continuous contact to share knowledge and co-design concepts. Prediction: Middle management should naturally embrace the concepts as meaning is created and circulated. Top management will sponsor the initiative due to emergent strategic interest.
Generative processes	Users and observation studies provided the foundations to generate concepts and solve existing or non-existing issues creating justified value-added. Co-design sessions with BUs also contributed to it in addition to knowledge sharing. Prediction: The gradual sense-making and demonstrated client value should ensure concept appropriation and development.
Environment cognition	BUs’ environment knowledge was integrated as well as extended by scouting. Prediction: The explored environment with new conceptual alternatives extends the network for BUs and allows them to build system thinking.
Organization design	The exploration project used internal ADT resources and progressively solicited BUs’ resources for exploration. Prediction: The proposed concepts may call for an intra-BU or inter-BU development depending on the interactions with existing BUs’ product architecture. BUs should then align as value is demonstrated.

Encapsulated model predictions for ADT/Design Thinking projects

Model descriptors	Predictions of the adaptive model
Model of coordination and collective action	The exploration projects were part of the overall ADT’s program of making boundary-spanning proposal for BUs to develop intra and/or inter BU products. Prediction: The sum of projects would potentially generate organizational learning for BUs, if not, create product development opportunities through top management program governance and sponsorship from VP of Business Development and Strategy.
Generative processes	The Design Thinking, combined with ADT’s experience and Practices, supported the exploration effort, in addition to the ADT overarching design brief. Prediction: The generativity would be driven by user observation and BUs segmentation to create the appropriate associations.
Environment cognition	The generative processes have encoded the need to scout the environment by targeting user experiences. Prediction: The exploration projects would open a new and demonstrated product development environment for BUs.
Organization design	The exploration projects rely on ADT’s resources and handover to BUs after selection. Prediction: Transferring would imply technical documentation to fit BUs available resources or identify new ones.

Anomaly: client value demonstrated but no NPD

Model descriptors	Actual course of action
Model of coordination and collective action	The exploration is guided by exploitation objectives (user pains) that are potentially in-between BUs boundaries.
Generative processes	The generativity is supported by Design Thinking and added practices (multi-BU co-design workshops). The assumption is made that the user and client value being demonstrated, the concepts will naturally trigger their product development. The generative process is naturally fixated by users (encoded in design method).
Environment cognition	User experiences and BUs’ environment are scouted. Design Thinking mainly develops concepts fixated by the existing environment. When crazy concepts are generated, they are adapted to what a BU would develop without considering capabilities (re-)generation of the environment (product standards, certification, sales channels).
Organization design	When crazy concepts are generated, they are adapted to what a BU can develop without actually considering new or recombination of capabilities within the BU or with other BUs.

The comparison of the actual course of action with the models’ prediction reveals again the necessity to understand the micro-foundations of exploration and exploitation, and their balance. The generative process practice of Design Thinking gives an idea of the direction of generativity and how it tries to build to innovation potential of attraction through demonstrated user and client value.

However, none of the business units were interested in actually committing to pushing the concepts further. They had accepted the concepts and recognized their interest and client value. They had even co-designed the concepts with ADT members, validated by VP Business Development sponsorship.

The exploration opened the path to new knowledge and better awareness of the environment and system thinking. Opportunities of connecting BUs together, through concepts revisiting interdependencies, were identified as they would be usually hidden behind their traditional market and product perspective. Nevertheless, the models fail in explaining why the concepts are not developed by BUs as we discuss below.

Co-design, acceptance and no follow-up

The non-invented here syndrome could be of course flagged as being the main reason for BUs not accepting to making these concepts their own. Indeed, no project lines were added to the budget by Business Unit. Only one added a line, but with zero budget, meaning they would perhaps consider a paper study. The death valley argument (Markham et al. 2010) is also not really valid as BU can budget the project, and budget consolidation is discussed with group CTO, ExCom and each BU manager to decide how they size the budget with support from the holding’s budget.

What is even more surprising is the fact that engineers, designers and managers from BUs participated to co-design workshops with other BUs. They legitimized and made sense of the concepts themselves in addition to welcoming the shared knowledge on users and value chain.

Exploring alternative aircraft spaces

Finally, the concepts generated with Design Thinking, as presented in our publication, were designed as if for BUs wouldn’t change. Moreover, the decision process to rank and select concepts in order to be promoted to BUs reinforced the stability of existing boundaries and engineering capabilities, as well as markets domains.

In some other cases, concepts generated through What Ifs? scenario allowed relieving some operational constraints (e.g. passengers storing their personal luggage in containers, passengers participating to waste sorting). These concepts unlocked interdependencies between aircraft equipment and engineering as seen in C-K reference mapping. However, the concepts were promoted as if BUs could just adapt them, without emphasizing the radicality of the new perspective implying deep reconfiguration of engineering design, system architecture and market positioning.

We have overall a good example of ambidextrous organization with a supportive top management leadership, a valuable generative process, deep interactions at different management levels. But, our literature models fail to explain the absence of further product development or even re-use of a feature in their technology and product roadmaps.

---

Chapter synthesis

---

References

---

1. It is almost a religious moment for engineers and the client who is sometimes invited. People come to witness the first dynamic crash test as it were a major milestone in the qualification process which validates design choices ↩︎