Chapter 11 - Intervention and steering committee

Keeping up with our research agenda, now that we have validated our model with two comparable case studies, we engage with the presentation of how the researcher intervened in the firm to test in vivo the model of decisional ambidexterity. It allows us to further specify the subtleties of the model and the management requirements revealed by the mutual conditioning between exploration and exploitation.

We show how the project trajectory was set and guided with design of decisions constantly articulated with fixations and interdependencies so that the engineering practice encapsulated in the project could be largely grounded and supported by stakeholders. Furthermore, we exhibit how this R&T funded project aiming for a potentially new product embedding TRL6 technologies, managed its organizational ties to find sufficient echoes inducing the regeneration of engineering capabilities enabling future new product development.

We come back on a management device we often referred to inter-BU cases such as the Design Thinking cases, the Lower Deck and the Connected Cabin. The Multi-BU committee, created in 2012 by a former CEO, was an opportunity to foster top-down projects as well as some bottom-up projects requiring strong sponsorship enabling corporate entrepreneurship.

It is not a case study comparable to others, as it cannot be identified as project on its own. However, it is rather a governance body for radical innovation. This locus of innovation within the conglomerate of SMEs can be analysed through the lens of decisional ambidexterity and we propose to specify some of its characteristics.

This chapter further tests our model of decisional ambidexterity and continues justifying its validity to extend models of ambidexterity and exploration project management presented in the literature review. These series of arguments provide us strong foundations to go into the following chapter where we present the researcher’s contribution to the quality management system for Design and Develop and Innovation processes (EN9100).

Hypoxia: probing and structuring the ecosystem interdependencies

The case presentation (see section) gave a first global picture of the project’s trajectory and context. The specificity of this case is that the researcher actively participated to the project on the R&T manager’s request.

Below, we propose to present what was specifically done to implement heuristics of decisional ambidexterity and its induced effects. We picture the trajectory with the global descriptors in Figure 2, as well as the phenomena relative to model of coordination and collective action, and its innovation potential of attraction.

Organizational context and project target

The Zodiac Oxygen Systems Europe (ZOSE), whose Engineering and Product Development board is common with its twin sister BU in the United States (ZOSU) is one of the two market leaders in pilot, cabin crew and passenger oxygen masks in addition to oxygen tanks. In the past decades, with ageing patents, an increase of replacement parts providers with an Parts Manufacturer Approval (PMA), the risk of obsolescence and market share losses became higher to the point that Engineering Director included it in its roadmap and practices the necessity to think of market regeneration.

One of ZOSE flagship product, EROS mask (see Figure 1 below) along with an equivalent product offered by the competition (B/E Aerospace), are provided to aircraft manufacturers on catalogue. They are issued with a TSO (Technical Standard Order) that allows to certify a piece of equipment with sufficient autonomy over the aircraft manufacturer. It induces then several design rules with associated qualification standards.

The equipment was originally developed upon transferred and licensed technologies from the US to support equipment design and supply for Dassault (Darrieulat 1993). Later, in the 1960s, the engineering department developed its own capabilities and winning its contracts on its own. In a pure entrepreneurial spirit, Georges Gutman, then development engineer, created a oxygen regulator solving several limitations such as having oxygen on demand, making breathing more comfortable.

Improvements were made including the innovation of having an expandable/retractable harness to adjust with the single push of a button the mask to any face whilst guaranteeing a perfect fit and air tightness. Smoke and fire protection with fogging prevention were also developed with a fixed or removable lens attached on the oro-nasal piece.

Given a historical and established dominant design, market players have put on the top of their priorities the urgency of reinforcing safety and preparing for the unexpected with new designs. Indeed, several threats of obsolescence due to ageing patents had been identified, and the aviation ecosystem was concerned by some isolated cases of hypoxia¹, usually due to other equipment failure. A CleanSky2 project, with Airbus as partner, made a call for proposal considering the new safety regulation requiring extended mask wear whenever a pilot is alone on duty in the cockpit.

It was then a proposal to rethink the ergonomics and features of the oxygen mask for this new restrictive requirement. Indeed, several feedback had pointed out that pilots were not following the rule because of the discomfort caused by the mask on their face. However, ensuring a pilot safety, whilst alone in a cockpit, is a first step before removing the copilot position. The flight engineer was removed several decades ago, so it appears to think of one-man cockpit before even examine unmanned commercial aircrafts.

Project trajectory

As explained in the case presentation (see section), the researcher had first proposed a C-K mapping of the Open Innovation Challenge in order to position the proposed concepts relatively to the dominant design. And, as he was familiar with the R&T manager who was first in charge of the project, he was requested to support their exploration and their methodology.

They had already been working for 6 months on the project. Several brainstorming sessions, workshops with industrial designers were held and digital mock-ups had been generated. However, they were confronted with a dilemma. They were expected to provide a product concept to Airbus: a product requirement had been issued largely inspired by the general pilot oxygen mask but with some changed words calling for innovative design. For instance, the word mask was replaced by device, the adjective disruptive had been added on some requirements, and they also referred to a future cockpit without clearly specifying its characteristics.

The researcher supported them in clustering their previously generated concepts, and in creating meaning around new product categories. The clustering was eased by the use of C-K theory. But, more importantly, the researcher insisted on the team’s ability to discuss the future decision of committing to a product design and development. So, the interactions with the client should explicitly rely on the extent of their design effort provided and their capabilities, as well as the necessity to discuss fixations and interdependencies. They were encouraged to present the design method, the clustering and how they made sense of the uncertainties and the unknown surrounding the ill-defined problem.

The figure 2 below shows the trajectory of the project. Initially, the concept could have been simply an extreme optimization ($d^+$) of the existing mask. It was a simple and easy way to identify action necessary for legacy products. It would allow anchoring to the dominant design and established product roadmap.

The necessity of wearing the mask for extended periods of time actually ensures that before an actual emergency, the pilot is wearing the mask in a standby mode, so the researcher pushed forward the nuance of prevention to counterbalance the emergency mode. It created a category discriminant to make sense of the generated concepts through brainstormings and industrial design.

Consequently, it allowed to consider two other concepts to decide that would fall under genericity ($d^{gen}$) and alternative situation ($d^{alt}$) as they wished for an evolving regulation where the nuance between prevention and emergency is specified. Finally, since the future cockpit is yet undefined, the potential threat of a guaranteed air tight cockpit would make oxygen mask pointless ($d^*$)? Furthermore, without fully exploring such disruption, they took into account nearby equipment that could be integrated with the oxygen device ($d^{alt}$).

All of the designed decisions where mapped through C-K with the project team input, and iteratively discussed so that they would gain in familiarity with the model. The researcher also encouraged to come up with means of avoiding isolation of the project within the engineering department. The project manager also repeatedly came to ask specific questions to the researcher at the beginning as the heuristics were not made clear enough; so some short sessions were quite randomly set at the beginning with him and his lead engineer, to give some guidance to think of the different decisions categories and the required engineering design.

In the following subsections, we offer more detail on the benefits of decisional ambidexterity and its practice through the researcher’s intervention.

A conservative client asking for radical innovation

Paradoxical request

We stated that the team was facing a dilemma when the researcher was called. The request made by the client was a mix of encouraged innovative design, but at the same time heavily constrained by mimetic requirements from existing dominant design. It created a lot of confusion and they were quite lost on ranking critical to quality requirements. They also doubted on the client for knowing what they were expecting from the product design.

The request was already ambidextrous in itself but perhaps in the wrong way as it was prescribed in an almost client-supplier relationship (Airbus to ZOSE). The team had the expectation that Airbus knew their need, and symmetrically ZOSE would engineer whatever is physically possible as long as it sustains and extends the product line. Reaching out for methodical innovative design, managed value management and project-based concern for engineering capabilities are clearly formulated despite being an open-ended exploration project funded by CleanSky2 program.

This is where decisional ambidexterity can help specifying the nuances between pure exploitation (prescription, engineering and development) and exploration. Decisional ambidexterity overcomes paradoxical tension of non-mutual conditioning.

Exploiting for exploration

From the perspective of C-K theory, having such specified dominant design makes it easier to build the rest of the concepts. So, the researcher did it whilst also re-using some of the mapping made for Open Innovation Challenge. The researcher then encouraged the project members to anchor the value of the generated concept to established requirement by specifying the distant and its nature.

It allowed to make a design document presenting three main concepts categories, with a digital mock-up each, stressing how they came about these options. The researcher had insisted on the necessity to explain their method, their line of thought and its decision parameters so that they could better discuss underlying hypotheses that Airbus may be deliberately or not hiding: interactions with other projects, cockpit elements, physical constraints, or just a vision for their future cockpit.

It has also forced Airbus to be more specific and answer questions that they did not expected to. For example, several debates gravitated around the head clearance (i.e. distance between the cockpit sealing and pilot’s head top) as some design features encouraged the use of surrounding space for prevention. The concepts considering the integration with other cockpit elements such as the seat (engineering knowledge and capability exist at Zodiac Seats France), forced Airbus to solicit FAA/EASA to start thinking carefully about potential evolution of regulations.

Opening the project routines to the organization

Stress-tested proof of concepts

The concept categories presented and the design method were really helpful in discussing the unknown fraction of the project, but also key to articulate the interdependencies of the future product and technologies. Due to the nature of the requirements demanded by the client, several interdependencies could have been easily challenged.

With the project manager’s support, the researcher encouraged the need to specify ripples caused by different generated features. The purpose was to find knowledge interactions and conjunctions triggered by the designed decisions. The manager organized open weekly meetings inviting several representatives from quality, purchasing, testing, prototyping, repair shop and other engineering sub-units. The aim of these sessions was to identify fixations held by these sub-organizations and the interdependent ties maintained by their engineering or management practices. Consequently, the project team was able to specify further the different concept categories.

Looking for organizational echoes

Nevertheless, the most stimulating side-effect of this practice was not the manoeuvre of the project trajectory to provide a well-defined and methodological approach to the client to co-explore the unknown. It was the reflexion of these practices in BU sub-organizations. The interactions organized around decision parameters, product design features and requirements were the occasion to regenerate their capabilities.

The Engineering Director launched several side mini-projects within some of his engineering units to address radical changes in product architecture. These have proven the technical feasibility of some features envisioned in the decision categories, and some opened new perspectives for product lines. For example, the air regulator has always been integrated in the oro-nasal piece because of pressure losses between the system and mouth/nose. Distancing the regulator through a hose was unimaginable however interactions during the open-meetings to update and consult BU stakeholders and prototyping allowed reconsidering its engineering and interfaces between sub-units. It also invited them to think of product systems and cockpit integration differently, and opened possibilities to change the engineering organization.

And finally, despite having been made fun of at the beginning of the project by other BU members, some of the originally generated concepts were indeed seen as coming out of the blue. These concepts were in the end grounded with exploitation regime and the distance created by exploration was sustained by knowledge identification, generation, and initiated organizational, generative learning and change. Decisional ambidexterity helped stimulating organizational metabolism.

Decisional ambidexterity analysis

MBU Committees: a steering committee looking for purpose

The Multi-BU committee was established by the ExCom to launch top-down projects as well as to offer a place for bottom-up initiatives. It was created in 2012 hosting to steer key projects whose strategic interest was acknowledged by the ExCom. For instance, the fuel cell development was initiated by the former CEO as the greener aircraft initiative became more and more important in the aviation sector. A project team was constituted and hosted by Zodiac Oxygen Systems Europe as a potential growth driver and technology transfers. The project aimed at developing the technology for aircraft use with the promise of having multiple sources to be reuse across the aircraft (electrical power, water and heating). The project was conducted in partnership with two aircraft manufacturers and funded by European programs.

The MBU committee also hosted numerous other projects that would call less for a new BU, but rather insist on inter-BU collaboration; others were rather intra-BU bearing a strategic interest as it was developing a game-changing technologies.

As we explain, in the case presentation (see section), the committee with its quarterly meetings was a locus of innovation to give a chance to entrepreneurial spirit².

Organizational context and ecosystem

First of all, it is important to understand that the MBU committee was set at a time when the homogenization of the conglomerate of SMEs was barely starting. The re-branding of business units was starting and the rate of mergers & acquisitions was diminishing.

The business units were exploiting their respective niches, and ZA as a group already covered a wide range of equipment and systems across the aircraft. The former CEO as well as other ExCom members, are all engineers and have a propensity for technical discussions. Niches were segregated with a variety of technologies, with high entry barriers. BU managers and the ExCom helped identifying M&A targets to be the first entrant to gain a strong foothold on these markets. The pattern had been reproduced across all existing ZA market segments.

The ExCom, as reported by the former CEO, wanted to leave a strong responsibility at each BU level, with Profits & Losses reporting, a certain financial autonomy with a holding consolidation. They avoided imposing projects to BUs. The MBU committee was then a means to drive exploration for BUs through technological development. Some technical barriers (engineering, certification, market interest, supply chain, etc.) would be extensively studied so that one or several BUs would take it onboard. If necessary they arbitrate on the project’s allocation depending on how the ExCom judges the BU capacity to sustain the development. The former CEO also saw the MBU as a management device to overcome the limited bandwidth of BU and Divisions managers. It is also the role of branch executives (ExCom members) to seize and support these initiatives.

Managing exploration added value for the conglomerate

The committee was a locus for discussion, where top management met with middle management, engineers, designers and marketers. All ExCom members were former engineers, including middle management and BU managers. As confirmed by the former CEO, he encouraged debate on the quality of models and input data used to make decisions on the technologies, products to be developed. The VP of Business Development and Strategy would usually question the Intellectual Property and aircraft manufacturers’ risk of claiming primacy. The former group CEO and branch CEOs would dig into the details of the project, questioning engineering topics, partnerships, competition, market shares, suppliers and potential launch customers.

Whenever they felt the model was stress-tested enough given their knowledge on the topic or insight they had from their interactions with other executives inside/outside the firm, they would delegate the responsibility to the BU.

Advantages, conditions and limitations - former CEO’s feedback to the researcher

Among the limitations we can find in the MBU device, some can be seen by reflecting on the difference between two cases such as the Connected Cabin and the Lower Deck. These two show that despite benefiting from the same context, they did not hold similar results. Of course we could go on listing numerous reasons, but they were treated alike during MBU and additional ExCom meetings. In other words, the same discussion on hypotheses, input data and modelling was heavily challenged by board members. Such pattern has been reported for executives not familiar with topics (Loch, Mähring, and Sommer 2017), but here, the ExCom is made exclusively of engineers who have cultivated over their careers a deep interest for technologies and applied sciences. They also have market and ecosystem knowledge through other firm’s corporate governance participation and external networks (consortia, group of industrial partners, conferences, etc.). For example, the former CEO had good knowledge for both topics before they were launched as actual projects:

• For the lower deck, he was aware of the decline of the air freight and its share in commercial cargo area. He was already convinced by the market interest and value for airline yield management, to the point that when the results of the market study were presented, he cut through the presentation to address technical and engineering issues. He would be more interested in securing the concept on that issue to gain autonomy from aircraft manufacturer and raise the entry barrier as they knew competition (B/E Aerospace) had been consulted for a similar project.

• For the connected cabin, he challenged a lot the value proposition as he was asking for proof of legitimacy for ZA to organize such service market platform. In an interview in May 2018, a year after the AIX 2017, he told the researcher he tried to encourage Zodiac Data Systems BU manager to develop a technological advantage securing the nascent market. It was another way of scouting for different scenarios and testing the fixations of different BUs as well as interdependent engineered product between BUs.

Despite having same discussion goals to define what makes rationality collectively with project representatives and potential stakeholders, it appears that the steering committee could benefit from a technology of organizing, by opposition to fostering a culture of curiosity and sound debate³ to sustain innovation.

It appears then that several conditions can only be surfaced through engineering design, project-based management and change management and its active management around a model such as decisional ambidexterity. Interactions and networking are not sufficient. Indeed, the former CEO recognized the interest of having a practice to specify the model testing, rationality testing practice on countless occasions. He was very receptive to the decisional ambidexterity model and even felt it was natural. The formalization is a real plus for a set of heuristics that may not be acknowledged by everyone. Otherwise, he believes that it is a widespread practice which counters procedural practices fixating rationality. He also insisted on the danger of such mechanistic models of thought (e.g. rational choice theory, operations research) which can be the best way to go to the wall.

Chapter synthesis: a reconciliatory model for ambidexterity

References