Bridge Design

These trends are exactly that trendy. Like any new method, they go out of fashion after a time. Sometimes it ’s for a good reason, but occasionally they are wrongfully discarded for the next new thing. Although many aspects of these trends have something to offer, the biggest improvement will come from taking a more-holistic view of innovation by focusing on three core areas of a business: culture, process, and resources. Try introducing six sigma into the wrong culture and it is bound to fail. Pour resources into a development program with a bad process and it is unlikely to create breakthrough products. This article provides insight into the aspects of company culture that make many successful innovations possible.

The trick is to experience failures quickly. Test ideas in the first months of the development project, way before any serious engineering has been done.  Use every trick in the book to visualize the product and get it in front of customers.  Show sketches, prototypes, and simple foam models. Cheat prototypes into existence by cobbling together existing products or hijacking technology from other industries.  Show early ideas to potential customers.  Get ideas out in the open for people to criticize, but do it as soon as possible.

Paradoxically, teams that learn to be good at failing quickly often learn to become better at succeeding quickly.  To encourage risk, and hence innovation, all team members must become comfortable with this paradox and adjust their expectations as a project moves from concept to refinement.  Taking risks early is inexpensive and potentially rewarding. Taking risks later in a project is much less desirable.  Then, the die has been cast, and linear, predictable behaviors are essential.  This takes almost superhuman management abilities, because unpredictable behaviors must be encouraged early on, but later, when the project progresses or employee reviews come along, the same manager must hold employees accountable to their management by objectives. Or the managers must account for the team’s progress to an upper management that is more skilled at counting beans than at growing them.

A project leader with great interpersonal skills as well as technical chops is key to a successful team. But people are not born this way—they are cultivated.  Along with the top leader, every significant project should have submanagers who are being mentored to take the leadership position on future programs.  Train for the soft people skills as well as the technical know how to cultivate a team.

Project leaders need to be supported by upper management with realistic budget and scheduling that has contingency built in. The entire team does not need to know exactly how much contingency there is; instead, the leader listens to each subteam’s needs and allots it some of the scarce resources.  Other team members must quickly learn the reasons for the allotment so that it is understood that there is sound logic behind it.

But even without a contingency, the schedule needs to be plausible. There is nothing less motivating to a team than shooting at a target that is hopelessly out of range. Obviously, there is never enough time to get every detail of a p roject perfect.

It is inevitable that a team leader will ask for a few miracles and, as the joke goes, these will take a little longer. But when the entire team understands the bigger project goals both technically and from a business perspective, and it is plausible that with a little extra effort these goals can be grasped, people rally to meet the objective.

But in subtle ways, such situations can come up all over an organization as a project unfolds. Sarcastic comments around the water cooler about nascent ideas or overheard phone conversations that poke fun at a part of a project that failed puts innovation in a straitjacket.

Experience often gets in the way of new thinking, but it is an important partner in making things actually happen. Therefore, the entire team needs to encourage a democracy of ideas, especially in the early concept stage.

How does a medical device company create a new product for a well-established market when a handful of competitors seem to have sewn it up? This was the problem facing Smiths Medical MD, (Minneapolis USA) who had recently successfully entered the diabetes market with an insulin pump and were looking to create a new infusion set for the same market. How could this new infusion set be made compelling enough to move customers entrenched with the competitive offerings over to a new product?

This article uses the case study of the design of this infusion set to illustrate how designers and engineers can best refine their design process to create new products that both benefit consumers and make their companies more competitive in the global marketplace. Although it is focused on a new medical device,the design process described here can be used across many product types that have a high degree of human interaction, ranging from consumer and automotive through a wide variety of industrial and scientific equipment markets, ensuring that a product is beneficial for users without sacrificing the company’s competitiveness.

Living with Diabetes

To understand this product challenge it is necessary to understand how patients currently manage their diabetes. Type 1 diabetes, also previously referred to as juvenile or child onset diabetes, is a condition where people stop producing their own insulin. Insulin is a hormone that controls humans’ ability to regulate the flow of glucose in their blood. If a person looses this control, their blood sugar oscillates wildly according to many factors including what they eat, how they exercise and if they are sick.  Such a lack of control over blood glucose leads to many life threatening complications such as loss of limbs and sight, and leads to shorter life expectancy for the millions of people with diabetes. Although much research is centered on finding one, there is currently no cure available. In the meantime people with diabetes have to control their blood sugar by taking regular doses of insulin, often in the form of multiple daily injections using simple hypodermic syringes.

What is an Insulin Infusion Set?

Increasingly, diabetes professionals and patients are recognizing the value of using a small programmable pump to infuse regular amounts of insulin directly.The programmability of the pump and the user’s ability to routinely adjust, or titrate, the dose of insulin allows people much greater control over their blood glucose levels. It also has the added advantage that users do not have to stick themselves multiple times with a needle. Instead they insert a small plastic tube, known as a cannula, into the midriff region of their bodies and let the insulin be pumped in exactly when needed. This connection, known as an infusion set, is typically worn 24 hours a day held on by an adhesive patch, and is disposable, with changes necessary about every 3 days. The cannula has to be introduced into the body with a small needle to help it reach the region of fat just beneath the skin (known as subcutaneous fat). After insertion the needle is withdrawn and disposed of.  A thin plastic tube then connects this infusion site with the pump.

As readers will see from the above explanation, diabetes is not much fun to live with. Patients have to suffer multiple insults to their body, including pricking their fingers for testing blood every few hours as well as sticking themselves to get the insulin in. Infusion sets and insulin pumps help, but as designers and engineers what can we do to make life easier for these people?

How can we make the new product compelling?

To really cater to a user, you have to understand exactly what it is like to live with the disease.  This is the premise with which Smiths approached the design of their infusion set. They asked the question “How could this new infusion set be made compelling enough to move customers entrenched with competitive offerings over to a new product?” Just doing a “me-too” product in the £XM diabetes infusion set market was not likely to win customers over or help the management at Smiths justify the significant R&D expense of a new offering.

Smiths immediately set about creating a small core team of about eight people to begin the development. This was a cross-functional team with mechanical and manufacturing engineers as well as product marketing people. Smiths medical also partnered with an industrial design firm, Bridge Design, at the very beginning of the project to ensure the team was rounded out and to take advantage of Bridge’s previous experience with developing Smiths market winning Insulin pump (Cozmo). The project started with all members on board and a team leader, Tim Bresina, who, although from a manufacturing engineering background, had a broad perspective on what it takes to develop successful medical devices and a real appreciation for what the various disciplines in his charge could do collectively to allow innovation to flourish. Team members often bring preconceptions or pet solutions to the problem they are trying to solve. This is inevitable, but before any serious work began on this project the team set about calibrating their understanding of two crucial things; what diabetics want and to benchmark the competitive landscape.

The “Deep Dive” into the Customers Mindset

To understand users, Smiths had three exploratory submarines in its fleet, all of which were important in taking a “deep dive” into the customer’s mindset. Firstly, its connection and understanding of its present customers was a valuable data point. Smiths had a nascent Diabetes business and a maturing understanding of its customers that were currently buying an OEM infusion set that Smiths branded and sold. The marketing people on the team normally worked in this marketplace, and were thus able to bring this perspective. Secondly, Smiths had assembled a couple of external advisory boards consisting of various types of healthcare professionals such as Endocrinologists (the medical doctor specialists for Diabetes) and Certified Diabetes Educators (CDEs) who are the nurse practitioners who work directly with patients to help them get better control of their blood glucose levels. Regular meetings with these advisory panels helped the team understand emerging trends, providing a forum for rapid feedback on product development ideas. With a patient’s purchasing decision influenced by a mix of doctor, CDE and peer referral, as well as the medical insurers or healthcare systems willingness to cover the cost of specific products, the market for infusion sets is complex. As Smiths was targeting a global market, they used ethnographic research to “get under the skin” of what people with diabetes really desired in the ideal infusion set –and this proved extremely important.  Ethnographic research is a technique of observing and interacting with people to gain a deep understanding of their needs, including needs that the users might have a hard time articulating. There are many ways of doing this research and they vary according the nature of the problem. Smiths enlisted the industrial designers from Bridge Design to do the ethnographic research for Cleo but also made a point of sending along a few of their engineers and marketing people to observe.  Bridge recruited about 50 people with diabetes of various backgrounds and ages all of whom were currently using infusion sets. To ensure a good spread of feedback three different geographic regions of the USA were chosen.

Once the participants were identified they were briefed about the team’s interest in better understanding what it is like to live with the current infusion sets on the market. Disposable cameras were sent out ahead of our interviews and two weeks later we met with them. Fortunately, our subjects were highly enthusiastic about it, and willingly told us what it is like to live with infusion sets, sharing photographs of their infusion routines and in some cases showing us where they were wearing them in the actual interview sessions. Of course, the interviewing process had to be conducted very carefully, with the risk that, if it was too structured, it could have prevented us from pursuing interesting lines of questioning. In practice it is best to have interviewers who bring very few preconceptions to the session but who have a significant understanding of the issues. For instance, participants will often tell little white lies about some of their routines because they know that they are “supposed to do it like this”.  In doing this research there are of course no right or wrong answers, so the role of the interviewer is to make subjects feel comfortable about being honest in their responses and to probe areas where there is concern that actual practice may differ from what is recommended.

Briefing the Team – Getting the Voice of the Customer into the product design process

A few core members of the team then set about preparing for an intense two day team briefing and brainstorming session where about eight team members from all the different disciplines would meet off-site for focused idea generating sessions. The focus of the sessions was not on solving a narrow set of technical issues. Instead it was on understanding the users’ needs for the product and using that insight to focus the whole group on a series of questions that were carefully crafted to elicit design ideas that attempted to meet those user needs. 

Prior to the two-day session, various team members also did some preliminary brainstorming around a sub-set of technical issues. The mechanical engineers from Smiths developed some interesting ideas for sprung mechanisms to assist with extracting the needle from the cannula after it is inserted through the skin.

Bridge Design went out and did some shopping for the project, spending about a day scouring the aisles of supermarkets, drug stores, toy emporiums etc. The designers found many interesting products that could be used in the brainstorming to stimulate the team.  Bridge design hosted the meeting in San Francisco and compiled a two-hour briefing presentation for the team. This briefing gathered all the research, including the ethnographies, benchmarking of competitive products and inputs from marketing and the advisory boards. All team members were encouraged to contribute their own experiences and research to the briefing–exposing the entire team to diverse opinions and interpretations of what the customers wanted. However, with a large patent estate already surrounding infusion technology, the team also had to consider what solutions would be out-of-bounds due to pre-existing intellectual property.

The team then set about debating what the customer requirements were. It is important to understand that users will not always articulate fully their hopes for what a new product might be like. People are very good at judging products and solutions that are already created; they are not good at imagining “What could be”. Therefore in debating the customer requirements the group was trying to get to the very basics of what customers want as well as what they need, which is why it was so important to involve a diverse group who had all had some kind of customer contact. In the case of a medical product, such as the Cleo, the customer is broader then just a typical user, as the opinions of healthcare and insurance professionals also need to be considered.

To organize and prioritize the list of requirements, which were deliberately stated very simply, a scoring system of 1, 3 or 9 points was used.  This scoring system was set up to polarize the scores around some big number differences as arguing about whether a feature merits a score of 4 or 5 does not force a team to make bold decisions about what really matters to the users.

The team also debated other factors such as the users’ current satisfaction with these types of products and the potential impact of meeting a particular requirement to improving sales.  For instance users expect a medical product to be reliable so this would have a neutral effect on sales, whereas creating a highly integrated device (insertion, adhesion, needle safety etc) would be a big plus for sales so is scored more highly. The purpose of the customer requirement ranking exercise was not to get the scoring ‘exactly right’. Instead it is an exercise that takes about two hours and is about the team debating,arguing and reaching an understanding of what these requirements actually mean to users.

This technique is a highly simplified adaptation of a system for product development known as QFD (Quality Function Deployment) that is often used in six-sigma types of product development environments. In practice this author feels that although QFD may be of value on very complex product development processes it is too cumbersome and frankly a very boring way of getting a development team excited about the product possibilities.  But stealing a few of its techniques and simplifying them are useful. (For a more detailed explanation of this brainstorming process see the article “Beyond Brainstorming” on http://www.devicelink.com/mddi/archive/04/09/013.html)

What the team learned from users was that they valued good old convenience and reduction in the steps to use the product. Children and some adults also expressed a dislike of handling the needles required to help insert the set. The requirements in this chart capture the essence of what the team felt customers wanted and importantly give some dimensions to the criteria rather than just blandly stating things like "convenient to use”. This is a much more useful place to launch into a creative brainstorming session than trying to plough through a detailed 50 page customer requirements document.

Setting up the Brainstorming Questions

Next the team set about creating a very pointed group of about six questions that would be used to focus the idea creation process.  This is a skilled task as the questions have to highlight the customer needs without limit or suggesting possible solutions.

The brainstorming itself consisted of about six half hour intense ideation sessions spread over nearly a whole day. Each mini-brainstorm was structured around a particular question centered on an essential customer requirement; ie “How can we reduce & speed steps in the insertion process?” The whole process of being immersed in the design issues off-site over a two-day period really focused the team’s efforts.  Keeping the actual sessions as short sharp bursts of creative energy with rest periods of at least ten minutes between sessions also helped keep the team fresh. Rather than using large white boards or cumbersome flip charts the group used small easily manageable “idea sheets” to sketch or write down their ideas and share them with the team before posting them on the wall. All the usual rules of brainstorming were applied and the team was encouraged to turn any critical thoughts they may have into new insights, building on both other team members’ ideas as well as generating new ones. 

One Idea a Minute

The result was that the team created about 200 idea sheets with innovative solutions focused by the careful preparation. The Smiths engineers shared mechanism ideas with the team and it further stimulated ways of using the mechanisms with a very user-centric appeal.  A smaller group of designers at Bridge Design then took the group’s efforts and sorted the ideas into three categories – Hot, Maybe and Not. As many readers are aware, successful brainstorming tends to focus on creating a large number of ideas rather than fewer higher quality suggestions.  This session produced a solid collection of ideas that were truly “Hot” with an equal number that were mediocre. Once the best ideas were sorted through it became clear that there were a number of innovative solutions that could be synthesized into different product concepts. 

The Whole is Greater than the sum of the parts

In all, about 6 different concepts were explored and the one that offered the most advantages was pursued and engineered by the Smiths team.  

The design attributes are:

• Users told us that step reduction was an important goal so the final product is the first all-inone system: sterile packaging, inserter and needle safe disposal container and reduces the insertion steps from about 15 to only 3. The adhesive patch does not require stripping as it is protected inside the disposable. The user simply unscrews the cap and gently presses the unit into their skin at a speed convenient to them. At the end of the insertion stroke, a spring returns the insertion needle up inside the container into a needle safe location.

• Cleo hides the needle from sight for ease of mind and perceived pain reduction.

• The “on-body” part was made smaller and adjustable to fit any infusion or pump location.

• Cleo is a discreet non-medical looking product that uses a classis bottle “rip-top” appearance to help cue users that this is a disposable product.

• It met the intellectual property goals of creating technology unique to Smiths.

You can view a short video of the insertion process here. You can find more consumer information about the Cleo here.  

Conclusions

Often engineering education focuses on the functional aspects of a product. In medical product design this is often thought of as how medically efficacious the product is. Unfortunately, in practice this view tends to disregard a broader user perspective – “How can you make this product easier for me to use”. In this context “easier” can mean many things.  New medical procedures often require users to change their habits, ergonomics and visual cues, instead of making a new procedure seem like a logical progression from something that they are already familiar with, which can really make adoption easier. An attractive product which appears non-threatening is also effective when trying to help children better manage diseases like diabetes.

The process that Smiths adopted for Cleo was really focused on getting these softer, user desired qualities into the product. It is possible to go back through the project history and see user comments that asked for certain attributes and then follow a trail through the design process to see how those insights really motivated the team to turn their technical skills in the most useful direction.  It is also possible to say that everyone on the team contributed to the design. Ideas were built upon; inspiring technical solutions came from the process. And, if a technical block was met along the path to implementation, the engineers at Smiths were highly motivated to overcome them as they had a good understanding of what their customers wanted, providing them with a sense of satisfaction of knowing that when they got the job done they would have made diabetes just that little bit easier to live with!

Left: Cleo with the cap removed and ready for insertion.  (Note that the adhesive requires no manual stripping.)  Right: After insertion with the needle safely retracted.  

What follows is an exposé of these issues and some practical insights into techniques that have worked in helping break the product development logjam.  We’ve also included some advice on what doesn’t work.

From this workshop group’s point of view, product definition falls into two distinct steps. The first is the gathering and documenting of the requirements.  The group expressed frustration that, in general, they are often insufficiently informed to their potential end-users’ needs and wants.  When data are presented to them, they are ambiguous, badly organized, poorly communicated, or all three. The second step of product definition concerns gathering user feedback once the product development process has begun.  The group said they believe customer feedback is not done often enough, and, when it is done, it lacks consistency and is poorly communicated back to the project team. 

These types of issues can result in a product that seriously fails to address its market, or a process that wastes resources and time.  The group had some positive ideas about how to rectify these problems.

It is unlikely that customers will consider every product feature before making a buying decision.  What will be your product’s big selling points? Define them early, then check with the market by showing customers simple mockups or cartoon storyboards. Conference participants suggested prioritizing features into simple classifications, such as “must have,” “nice to have,” and “acceptable.”

As the project progresses, keep reminding the development team of the top customer requirements, and use them to judge the product’s progress. They can also be used to help judge the value of competing solutions to product development setbacks.

Managing Resources for Timeliness

If you have never heard the phrase, “We could do it if only we had more money/people/time,” then you are not—nor have you spent time with—a manager or engineer responsible for product development.  We spent a half-day with this workshop group at the conference, and, as one might predict, resourcing was brought up as a major impediment to developing new medical devices while meeting business goals.

Something’s Got to Give  The three-way tug-of-war inherent in the product development world, often referred to as the “triple constraint,” takes place among the competing demands of quality, cost, and time. 

Quality also refers to scope. In our experience, discrepancies often occur in the way that management, customers, and engineers define quality for a given product. For instance, if a “nice to have” feature is dropped because of cost or deadline constraints, engineers often perceive that the product’s quality has been compromised.

Cost relates directly to the resources made available for the project in the forms of money, people, technology, materials, and outsourcing. 

Time seems always to be in short supply. Projects start too late to meet their set completion dates. The pressure can come from needing to have a product ready for an annual trade show, from the presence of a competitor lurking nearby, from a customer who requires delivery by a certain time, or from a management team whose annual bonus is at risk.  The old adage “time is money”is evident when it comes to developing a product.

Conventional wisdom claims it is always possible to meet two of the three demands but generally impossible to meet all three. And, in fact, workshop participants said, “We typically do not get enough R&D time to develop a refined product,” and “Funding resources are limited: scaling and objectives need careful clarification.” In other words, “Unless you give us more time, we cannot achieve desired quality,” and “Limited money means limited quality, or scope.”

The group identified several resourcing solutions that have worked for their companies, as well as a few that haven’t. In addition to providing these solutions, it is important to identify the assumptions that make the successful solutions viable, and those that fail, unworkable.

Which Is Most Important: Faster, Cheaper, or Better? If you can’t have all three, then which are most important? The solutions that translate into highly competitive, successful product launches are those that place time and quality at the top of the list (see Figure 2). The companies that do so recognize that being first to market means more than saving a few pennies or even tens of thousands of dollars. These are not firms that are putting inventory into their pipelines too early, or starting more projects than they have the resources to handle. Rather, they are companies that invest in the capacity required to develop and launch high-quality products fast.

Teams behind successful product launches also understand that uncertainty and variability are inherent in a product development environment. They make sure their plans do not commit more than 70% of their capacity and that those plans utilize time buffers. 

It is wise to identify tasks and parts that are high risk and make explicit contingency plans just in case the worst scenario happens. This includes maintaining a list of on-call outside resources. Ongoing communication with these contractors helps them to anticipate what assignments may be coming and prepare to allocate the resources necessary to perform for the work.

Finally, learn from past projects. Put in place an auditing system so that resource needs can be forecast clearly and improvement projects can be targeted to those tasks, parts, and resources that are most variable or that will require the most time.

Resourcing: Taking Variability into Account. One comment made about resourcing during the workshop was this: “On paper [a resource plan often] looks fine, when in reality it’s not—for instance, allocating 20% of someone’s time as a resource.” Another was, “Resource allocation is good for the long term, but it doesn’t mean we get things when we need them.”

These statements illustrate what happens when organizations fail to take variability into account. Developing—and attempting to control—project plans that assume infinite resources and specific start/stop times for tasks will likely result in unmet goals. The same can be said of plans started regardless of the capacity requirements placed by other projects, and plans that have not strategically provided for variability.

Nurturing a Supportive Corporate Culture and Leadership Style
Show Me How You’ll Measure Me, and Watch How I’ll Behave. The effects of corporate culture and leadership style on an organization’s ability to achieve successful product launches constituted the third major focus of the workshop discussion. One common perception expressed by the workshop participants is that many people resist new approaches or ideas. Another is that senior staff do not follow through on their commitments. Accountability seemed to be the word of the day. Participants tended to believe that “if only those senior staff members were held accountable, all would be well in our product development world.”

Nurturing the culture seems easy if one assumes that the organization is made up of adults who can and should be trusted and respected. Good communication and information flow are essential. And for those things, you need trust and respect. It should be noted that team members and managers need to communicate in more ways than by e-mail alone! Believe it or not, the statement at the top of the workshop participants’ “doesn’t work” list was “e-mail hell.”

Take Risks and Find Errors Quickly. The wise advice not to punish people for taking risks and making mistakes goes hand in hand with encouraging cross-functional team input and buy-in to the project plan. When the whole team offers input, more risks can be identified up front, more assumptions can be examined, and, quite often, alternative solutions that reduce risk can be found. These methods will work best in an environment in which all assumptions are
considered challengable.

In such an environment, real core needs are identified. For instance, a visual prototype may be acceptable for the investor review coming up in three months,but a working prototype will be needed for the trade show that’s six months away. When project plans are developed, real communication should take place between the project manager and senior management regarding any trade-off issues. Free-flowing communication is preferable to the typical “I need more time or resources” requests by the project manager being countered with the typical “We don’t have more time or more money, but make it happen anyway” response by senior management.

Don’t cross your fingers and anticipate lots of overtime and weekend work to make deadlines. Use the planning tools available in numerous software packages—in conjunction with the tool you’ve got between your ears. Share the results with the team leadership to help them juggle resource and schedule commitments. And when a project must be stopped, communication should flow to those people who have worked on it, so that they can understand the business reasons that called for pulling the plug.

Enlist an Experienced Project Leader.  The vast majority of the workshop group said that one of the most important requirements for a successful product is a single, experienced project leader. You need someone who understands the product, the customer, and the organization. This individual should know how to make things happen, smooth ruffled feathers, and see the big picture as well as the details.

This leader must be able to drum up creative solutions and distinguish “noise” from real problems. Not only should every project have such a leader, but every leader should be actively mentoring at least one team member to take the leadership role on the next project. Companies that are aware of the importance of strong leadership and consciously nurturing potential new managers are not only better places to work—they tend to create more successful 
products as well.

Finally, the measures that are used to judge whether individuals and teams are doing a good job must be in alignment with the overall project plan. As an example, if project tasks are to move from one group to the next without stopping, then don’t measure the groups on “task due dates,” and don’t treat the task estimation process like a typical budget negotiation.  Instead, measure the groups on achievements such as project success, project innovation, and development of the knowledge and skills needed by the organization.

The Common Threads of Success

It is clear, based on the feedback from this group of product developers, that team members are not alone in the typical issues they face when charged with bringing a new product to market successfully. The keys to success in product development are similar to those of the other great challenges in life, such as marriage or parenting. They are:

• Communication.
• Leadership.
• Monitoring what works and what doesn’t.
• Taking risks and learning from mistakes.

Despite having about 50 years’ collective project experience, one of us from a technical product design background and the other from a business-process design background, we both learned new things from this lively group of workshop participants. We expect to continue learning. Perhaps project managers should be even more conscientious about sharing their experiences—warts and all—with their peer groups. We know that the ever-changing world will never let us rest, and we’ll finish this next project and then turn around and do another one—only better next time.