Ian Nimmo on the changing DCS

Ian Nimmo is the expert on HMI and Abnormal Situation Management in the world. His talk this afternoon at the Expertune/TiPS User Group meeting is sure to cause some serious waves, as he predicts the demise of the major automation vendors: ------------------------------------------------------- The Rise and Fall of the DCS The DCS came on the scene as the biggest and most technological development in the last 50 years. We moved away from charts and trending. We used to be able to get a good picture of the plant by looking down the controller wall. The DCS continued to eveolve with additional functionality. PID was beginning to be replaced by MVC. Operators started to work with digital numbers instead of looking at trending. The DCS installation has become an IT nightmare as the DCS is now more an open environment. Networks and network management techs are as prevalent as instrumentation techs were in the old days. Firewalls and servers are almost as common as controllers. Remote I/O and Remote Servers with KVMs have left only a mouse, keyboard and screens in the control center. IT doesn't matter! Nicholas Carr said in 2003 that IT is very important, in the same way that electricity is important. Once avaiable to all firms, it becomes a commodity or factor of production, and it can't be used to differentiate or generate competitive advantage. It is therefore a part of the infrastructure. Let's look at the evolution of infrastructural technologies. We go through a proprietary phase after the idea emerges, then to a fluid state where competing ideas flourish, and then we see a dominant design where standards norms and best practices emerge. Finally, we get to a build out phase, and then to commoditization. On the supply side, we see a small number of suppliers dominating the early market, but later a large number of market entrants as the dominant design stage is reached. Huge market growth leads to universal proprietary standards or open standards emerge, until finally the components become fully interchangeable, outsourced and generic. On the customer side, early adopters collaborate with providers to achieve modest competitive advantage. Then users deevelop radical new applications, and significant cometitive advantage. This often leads to conflict with the suppliers. Finally the technology matures. In a mature technology, generic approaches dominate. There is rapid price deflation. Technological advantatges of suppliers are eroded, but they can be converted into positional advantage. The risk of downtime/loss of availability far outweighs advantages offered by incremental investment. Big outlays continue, but as part of overhead costs. Integration, inter-operaability and pervasiveness replace local optimization as the source of perceived competitive advantage. Where is Process Control? The buildout phase is coming to an end. Next generation DCS will be open standard with open and interchangeable components. Multivariable Control is entering the buildout phase. REaltime Optimization is still proprietary and may always be. Real time data exploitation in Fluid Phase. What will happen next? DCS platforms will become increasingly open. They will cease to exist as distinct entities. Instead we will build RT networks, selecting between components from multiple vendors. Open platforms will enable new phase of APC applications. There will be novel applications exploiting the abundance of realtime data. Traditional MVC will be integraed into the infrastructure. This will mean the death of PID and the real technological paradigm shift. STandardization of engineering functions such as alarms and the HMI into almost clip art like objects. What are the implications for Users? There will be an emphasis on openness of systems over capability and value for money. We'll delay expenditure until new products are established and proven. We'll initiate collaborative development of novel data exploitation applications. Purchase of services will be attractive over capital investment in hardware and projects. Control and systems engineering will be increasingly outsourced. We'll seek competitive advantage for region/sector through pervasive application of technology rather than for individual firm. And for the suppliers the implications are also grave. Radical adjustments of pricing mechnaisms and profit margins mean new business models. Software/systems integration/consultancy and outsourcing become dominant rpofit centers. Current distinctiveness and value adding capabilities will become liabilities in the future. You will need to convert technological advantage into positional advantage. Radical chagnes in user attitudes towards process control will force a significant restructuring of the supply base over the next ten years. Suppliers will be able to gain short term advantage through restrictive practices. Long term advantage will fall to those suppliers who embrace open standards early. What does this mean for functionality and standardization? The vendors will be reluctant to add functionality. Customers and users will have to adopt standards to ensure plug and play components and provide common look and feel. International standards will emerge due to continued growth of accidents caused by human error. The Honeywell or Emerson, or ABB or Yokogawa way will become extinct. There will be one alarm management system and one HMI standard. Human fators will dicate the vision for the change. Alarm philosophy and design guidelines will force conformance. There will be limited priority levels. There will be common ways to determine priorities. Common alarm functions such as inhibit, disable, eclipse, suppress; and common filtering techniques. Management of Change and documentation will become essential. This is a healthy future for TiPS and other enabling technologies. Human factors will determine what is acceptable. HMI style will become as common as Microsoft's Style Guide for Windows, dictating look and feel. Object libraries and dynamos will be as common as clip art and independent of platform, adn the numbe rof screens and navigation will be formalized based on the number of operators and the process. Common tools will allow configuration to any system, and large screens will become as common as the old mimic panels. In the transition away from the DCS, customers will migrate to contractors. Old practices will be followed. Everyone will do what is right in his own eyes. We will create a mess which will be costly to correct. We will learn from our mistakes. The wise will anticipate the future, and will develop an automation vision, and then invest in HMI Style Guides, Good Industrial Practices for alarm management-- these things are known and people are already implementing this. However, be careful. Don't get lost in poorly conceived ideas.