Regarding automation control and robotics efforts today, nearly 60% of all costs are due to engineering and commissioning. While in the past optimization strategies have addressed the bought-in parts well, most room to increase efficiency is now assumed in the field of engineering (s. Figure 1 below).

Source: Cost structure analysis of controls and robotics (AIDA 2005)

Figure 1: Analysis of Engineering Costs

Examples are media discontinuity beginning in plant planning, where hall plans (e.g. from microstation) are manually redrawn in another tool for line planning, over CAD development with a heterogeneous tool landscape over all suppliers up to logics development, where mechanics engineers develop time plans in Microsoft Office tools, that are only verbally re-used for PLC programming.

With the AutomationML™ as open intermediate format, we see many opportunities to improve automation engineering with new applications for better and more flexible collaboration as well as for earlier simulation and quality assurance, which will be based on improved tool chains. Each single aspect itself will lead to higher data and planning quality and reduced costs.

The line engineering will use the lossless data format to enrich the system specification step by step supporting early validation involving all disciplines, i.e. mechanical engineering, electrical design and control. All data can be tagged with version and author information. As a long-term goal, a complete virtual engineering is achievable combining e.g. CAD data, simulation, process data and controls.

The AutomationML™ as accepted intermediate data format will allow to improve the tool chain dramatically, as tooling will be far better taylored for each individual task. While the usage of expensive tools will still be appropriate for detailed design efforts, the communication between developer and customer could be supported by web based interactive viewers, avoiding traveling. The communication could even come for free based on open viewers already available today.

To be usable in the outlined engineering chain the only prerequisite for a tool is an AutomationML™ connector, i.e. an importer and exporter. Demonstrators and concrete usage of the current AutomationML™ version together with well established tools already have proven the basic feasibility of this approach. The underlying technology, based on the open standards XML as well as COLLADA for visual representation, is already field proven in multi billion businesses.

A very positive consequence of this approach thus is the strongly improved collaboration opportunity of the engineering partners, while their flexibility to choose their own tool chain is increased at the same time.

Tool vendors will benefit from focusing on their core competencies, because now, they only need to develop one connector to AutomationML™ instead of multiple for all available tools, and now they can potentially better use plug-ins from other vendors for certain useful but not core features, e.g. viewers, rendering, simulation engines, web collaboration tools etc.

Furthermore the reduced tool costs and introduction barriers for small companies as well as for research facilities and universities will allow broad innovation power and knowledge availability.

The most important aspects are improvement of quality and cost reduction. Quality will be directly addressed through usage of a lossless data format for the whole engineering process, leading to data stability across tools avoiding media discontinuity. We also assume that the seamless usage of results of preceding phases will lead to increased engineering quality. But central for quality improvement will be the far earlier testing, that will be possible with early and interdisciplinary simulations. Earlier quality observations will also be key to reduce planning risks.

Cost reduction will come in several ways. First and most important through quality improvement as described above, but also through reduced engineering costs and time by better usage of results of preceding phases instead of re-developing these results with other tools and formats.