Increasingly companies become global. They outsource some of their operations. They collaborate in product development. A capability to do that worldwide in product design and engineering becomes a competitive advantage. But how can this be achieved?
Three things are needed, the capability to share the design specifications of the product worldwide, the global management of engineering change and the capability to interact on the same data from globally distributed locations. Information technologies exist to achieve each of the three, but to date they have not been integrated closely to develop that integrated approach.
Companies have different collaborative design approaches, going from designing the product internally and outsourcing its manufacturing to OEM’ing a product designed by a supplier. We recognize leveraged, joined and directed design, as intermediate steps. Each of those three require a form of collaboration between partners that are often located in different continents, making collaboration difficult. Web 2.0 approaches should help us improve the experience of such designs.
A from of cloud computing can serve as the basis for the sharing of design specifications. I believe that large companies will focus on the use of a form of private cloud in the short term, as security and IP protection issues may keep them from exposing their future product information on the public cloud. Unfortunately, when using CAD/CAE tools in particular, the network latency typically becomes an issue. Two approaches can be taken to address this problem. The first is to locate around the globa a number of staging servers, where a copy of the files are stored to improve responsiveness of the system. The master file, located in the cloud, is then updated in the background. by transferring the changes. This works well for leveraged and directed design where one of the parties performs the main design tasks. In joint design, things become more complicated. Both parties need to be able to work on the same designs concurrently. One way of doing this is by maintaining the CAD/CAE functionality in the cloud and use remote graphics features to transfer the screen content to the geographies. This reduces the information that has to be transferred drastically, providing an appropriate latency in most parts of the world.
Such approach is obtained by locating blade workstations in the cloud. These become the visualization stations, who transfer the pixel information for each pixel on the screen. A side advantage of such approach is that the user only receives pixels, and not the actual designs. This is a great advantage when collaborating with companies in countries where IP is not well protected.
Cloud based workstations are not usual with current service providers. This approach will require, at least in the short term, more of a private cloud approach. That does not mean the OEM has to manage the infrastructure, this can be done by a service provider, but it may have implications on the elasticity of the cloud. For this reasons, companies may have to stage their development cycles to ensure a reasonably constant use of their CAD/CAE environment, in the same way they do it today for their current environment, It might actually make sense for multiple companies in a specific industry, to group their needs and develop a joint PD&E cloud for their needs. As they often use the same suppliers, this would facilitate the adoption by the suppliers.
The one question that remains open is who pays the service? Is it the OEM or do all parties cover their own costs. In the first case we have a true private cloud, in the second, more of a community cloud. The possible business models start from OEM funding to subscription and pay-per-use.
I’ll address the other two aspects in my next post. Tell me what you think about the above?