Companies are exploiting all the power of those technologies to gain the competitive edge that is so critical to their survival - or are they? Many companies have failed to realise the full commercial benefits of these tools. However, PA Consulting Group's recent development of a breath-actuated inhaler has demonstrated how time to market can be reduced to a third of what would be regarded as the industry norm.
When PA embarked on the development of a new breath-actuated inhaler, best practice use of computer-aided engineering and rapid prototyping techniques were core to the development plan. What PA regards as best practice has been developed from the experience of numerous projects in the drug delivery sector and from the other areas in which it operates.
The key to success of projects of this type is generally to get the whole team to work together effectively. In common with many of PA activities this project involved close collaboration with external partners, so the project team would incorporate individuals from different companies, disciplines, cultures, and counties. All of these people would have to work together efficiently for the project to achieve its objectives in the time scales demanded. There are three main aspects where the technology available can facilitate the operation of the team. These are design communication, early virtual and physical testing, and effective management of project data.
Communication is so much easier if everybody can see the virtual object under discussion
Very rarely does a small team do drug delivery development on its own. A team will typically include analysts, marketers, industrial designers, engineers, pharmacists and researchers,. Rarely are the people with these diverse skill sets based in the same place; frequently, they do not even speak the same language. Good communication does not come naturally. The good news is that all of these people can now have a common language, "the virtual model", and the great news is that everybody sees the same thing, no special training is required to understand what you see, and everybody can use the same model for their specific task.
The virtual techniques needed to model the device, work on it, analyse it, and understand it - like three-dimensional CAD modelling, parametric modelling, and structural and kinematic analysis - have been around for some time. The techniques related to rapid prototyping - like SLA, resin casting, CNC machining, and injection moulding - are not new either, although some of them have developed rapidly over the last five years or so. What is new is that now these techniques have matured to the point that they readily work together, all will now feed from or add to the one underlying data set. Design changes can be quickly and easily communicated to the whole team; gone are the problems of human error in the interpretation of drawings, remodelling ideas for different systems, and some parts of the team working on obsolete versions. Suddenly, the nightmare of communication (or miscommunication) no longer seems so scary.
Figure 1 shows the skeleton of a simple mechanism for pumping fluid.
Figure 1: An initial idea is captured in the three-dimensional CAD environment as a two-dimensional line drawing
The design team will want to use different packages for different aspects of the design. The time savings that come from each package receiving data electronically from the one master data set are obvious but the benefits of accuracy and reduced errors are just as significant on the progress of a project. The images shown in Figure 2 are all based on a single data set from the three-dimensional CAD model.
Figure 2: Design visualisation based on a single data set is a powerful communication tool
Early testing (both virtual and physical) avoids costly and time-consuming errors
Errors identified in the course of engineering development are costly, and the later in the process they are discovered, the more costly they become. CAD systems make it feasible to move aspects of the testing process, both virtual and physical, way up the development chain. If virtual testing can be used to identify problems sooner the device will be successful in the market earlier and with a lower development cost.
Virtual testing can verify most elements of the design
Virtual testing can take many forms, but what all of these forms have in common is the ability to identify problems early when they can be corrected at minimum cost and with minimum lost time.
The most basic forms of virtual testing are accomplished simply by adopting the best CAD modelling techniques. Figure 3 illustrates a CAD model for the breath-actuated inhaler that not only looks like the product but also works like it.
Figure 3: Modelling the behaviour of the mechanism can virtually test geometry, interference checking and tolerance sensitivity
This model can be used to check the parts fit together, that the components can be assembled, the mechanism geometry is correct, that there are no interferences etc. By adjusting the dimensions of key components, the tolerance sensitivity of the design can also be established. While this is frequently not recognised as 'testing', that is, in fact, exactly what it is.
Physical testing proves functionality and performance, and assesses user interaction
Whilst new technologies have created new opportunities for virtual testing, there are also new methods to enable physical testing. It is now faster and easier than ever before to create physical parts directly from the CAD models. The project team used a very wide range of rapid prototyping techniques for developing mechanism, component performance testing, ergonomics, and crucially to engage doctors and patients in the design process and to obtain critical market feedback.
Best practice device development should always include testing design concepts with real patients to avoid committing significant work to the wrong design. Using today's rapid prototyping technology the project team progressed from designer sketches to design concepts in just five weeks, and was ready to start the market research.
Figure 4: High quality fully functional models produced in weeks rather than the months usually required.
Managing the data in a structured way ensures that all design input is consistently captured and easily accessible
In a project of this type, the team will generate and store hundreds and hundreds of electronic files, and depending on how this data is managed they can become either a nightmare of administration or the lifeblood of the project. It is not difficult to envisage the chaos that would very quickly occur if people on the project were to do things in an unstructured way.
Correct management of the project data is vital to the effective running of the project, most people will have their favourite "war stories" that relate to parts of a project team working on the wrong version of a design for months on end without knowing it. The underlying principle is to maintain one master data set that defines at any point in time the product the team is working on. At PA the concepts and procedures for product development have been designed and improved over many years and are based on the practical experience of a number of industry sectors. The procedures required to control the single data set are defined in the companies quality system and include aspects such as: configuration management, document numbering, part numbering and revision escalation, CAD filename conventions, change control, etc.
The ideas presented here are based around the improved use of what is well-established technology. PA is now actively and enthusiastically looking at the emerging generation of new collaboration tools. These products offer tremendous promise to again revolutionise and improve the way we work, but it is easy to be seduced by the technology and we should remember that the CAD industry has not always delivered the promised benefits in the past. Collaboration tools will only be of value if we can achieve further improvements in product quality, development time or cost, and for this the initiative is likely to have come from the development practitioners.
Of course, being world-class at developing drug-delivery devices requires investment in the best people and the best development tools, but it is knowing how to apply these resources tellingly to the development process that will deliver a real step-change in performance. Using the latest generation of CAD and prototyping techniques in a completely integrated manner means that communication is hugely enhanced; the fact that images and models can be produced directly from the CAD file greatly increases the range of people who can be engaged in the design process, while simultaneously increasing the number of tasks that can take place concurrently, thus reducing time to market. Both virtual and physical testing can be carried out far earlier and much more cost-effectively than ever before. Creating a single design- data file as a by-product of the process means that time and errors plummet because the same data is being used for every process. If everyone were applying best practice in this way, everyone would be designing, developing, building and testing world-class drug-delivery devices ready for clinical trials within a year. There is no black art at work here: it is within everyone's power to do just that.
