This post was originally posted on the CloudSource blog on February 8th, 2015
In many industries, Research & Development is the function that ensures the long term viability of the organization- through a continuous delivery of new products and services. In electronic consumer goods enterprises where average product lifecycle is counted in months, being able to ensure a continuous flow of improvements and new features is needed to keep products flowing.
Furthermore, as electronic consumer products features are increasingly implemented in software, delivering the software components reliably and on-time becomes critical. Actually, the same applies to the car industry, with one caveat that many car engineers are great specialists in mechanics, not in software.
The question then becomes: how can this be achieved best ? Being early in the market with a new functionality typically allows you to ask higher prices and to take market share. So, your goal should be to set-up a R&D environment in which your developers can focus on the task at hand, deliver new functionality fast and with high levels of quality. Collaboration and re-use are two key elements in making this happen.
R&D engineers are asked to be creative, to think out of the box. Following processes and depend on others is often foreign to them as it gives them the impression it limits their creativity. And they have a point. Until now, many organizations have coped with this, resulting in R&D using a patchwork of systems and software, making their management and back-up extremely difficult and ultimately putting the company at risk in many situations. So, what should we do?
First, Optimize end-to-end Engineering Capabilities
To address the challenges described above, you may want to transform your R&D approach. This requires a top-down approach, which can be a little tricky given creative resources often do not like too many boundaries. Thus, sharing a compelling vision is critical. You will need their buy-in, or at least the buy-in from a small team. They will become your champions and help you sell the new approach to the remainder of the organization.
Once the initial team is identified, sit down with them and strategize on how to improve. Look at the processes, the information available (component libraries, source code management systems, documentation sources etc.), the tools and infrastructure used and the administrative procedures. Take a lean approach, identify waste and remove it. That will give you the roadmap of how you will get from where you are to a more effective environment.
At that point, you will be able to start small working groups that identify how the processes will be redesigned to be more effective. If you have any 6 Sigma black belts in your organization, assign at least one to this working group so he/she can help maintain lean thinking. Also, take a close look at the agile development methodologies such as Agile Scrum. They will provide ideas not just for software development, but also on how to make your hardware developments more agile.
As this process continues, the team will identify tools required to do the job. Launch extra working teams that focus on the tool selection. And last, but not least, you may want to look at the organization as a whole in order to see if you can adjust structure, approach R&D holistically and speed-up the time to market.
Second, establish a set of integrated tools
While you defined the new way of working, you identified potential tools. The features of the individual tools are important, but the ease by which they can be integrated is critical. We want to make sure the tools can be integrated to minimize manual work required to update the different components. Also, ensuring all groups work with the same tools, so information becomes available to all, should be examined very seriously. This would maximize re-use of existing components and designs, further reducing the time it takes for R&D engineers to come develop new products while reducing development and production costs.
Tools should include CAD/CAE tools for mechanical and electronic components, software development tools, PLM and Master Data Management and simulation tools. Such integrated set of engineering and collaboration tools enables end to end optimization of key engineering capabilities.
I would suggest we add an internal social media tool to this. Such tool allows the R&D teams to collaborate more easily, using the community as a sounding board for ideas and as a resource for solving problems. Particularly with large, multi-national teams, this becomes important. One element close to my heart, is to complement this social media tool with a knowledge management repository, as you never know, other engineers may have the same or similar questions at a later stage. Capturing the information exchanged in a structured manner may allow them to easily find a response without having to go back involving the whole community.
In the process, capturing the skills and expertise of the individuals provide a repository of experts. They become the go-to persons, in case specific aspects need to be addressed. It can be very rewarding to be seen as an expert in a particular field. Thus, it motivates the teams to use the systems to collaborate, which is one of your goals.
Third, look at the supporting infrastructure
Standardizing the processes and supporting tools, centralizing the information changes the way the R&D engineers should get access to IT. Having the IT tools under the desk of each engineer, managed by himself, is no longer possible. The use of a central IT environment is required, and here is where cloud actually comes in. Providing engineers with an IaaS service in which he can provision the infrastructure, middleware and applications he needs at any moment in time. Such service can also include high performance computing farms enabling digital testing and simulation.
When it comes to CAD and CAE software, because of the need to manipulate large graphical files, the location of the R&D team becomes important. Going for one centralized compute environment may not seem appropriate due to latency and bandwidth limitations. There are actually ways around that. First, the use of virtual engineering stations and other visualization techniques may help bridge the distance. One drastic approach is to set-up an engineering workstation in the datacenter right by the cloud environment and only transfer the screen pixels. A side benefit of this is that, when you work with subcontractors in sensitive countries, you limit the capability to have IP leakage.
Alternatively, systems may be positioned in closets closer to the actual locations where engineers are operating. But what is important then is to make sure there is ONE master copy and that the master copy is centralized so it can be accessed by all. Copies should be managed hierarchically.
What value does such approach add?
At HP we went through such journey a couple years ago and let me share with you what benefits we got from doing this. First, we eliminated one design iteration, which obviously allows us to get our products to market faster. We recovered 10 to 20% engineering time as they did no longer have to manage and reconfigure their systems. The environments they need are available from the catalog available through the portal. The number of versions of middleware and tools used reduced drastically, facilitating support. The tool licensing cost reduced by about 40%, while the automation lead to 25% les admin costs. We used to have 68 parts library and consolidated those to 1. As a result, the parts library consistency went from 80% to 99%. And last but not least, our engineers now have 2.5 times the performance available with half the number of servers running on a quarter of the power and cooling.
Do you think, this is worth doing in your company? We can help you there. Interested in knowing what our vision is and how we can help you? You may want to read here.