<strong>Cyber</strong> <strong>Physical</strong> <strong>Systems</strong> – <strong>Situation</strong> <strong>Analysis</strong>DRAFT – March 9, 2012automation systems is that they can be more reconfigurable than custom-designed systems, but modulesthat have embedded controllers are more complex than a typical system and requires additional testingand validation before it can be applied. 39A study from the University of Michigan identified the following research needs in the areas of modelingand verification of mechanical and computing systems (direct quote from report): 40• Automatically generating models of physical systems that can interface with models ofcomputing system operation• Specifications of model correctness and uncertainty, and methods for validating thesespecifications• Abstraction methods for modules, to manage complexity• Specifications for re-configurability, and methods to validate these specifications• Specifications (interface models) that guarantee interoperability• Certification methods to guarantee conformance with international standards• Verification techniques for modular CPSThe study also identified a possible roadmap to achieve these needs, describing steps in specific order: 411. Survey of existing techniques for modular design in mechanical and software domains2. Mapping potential synergies from mechanical and software design domains to cyber-physicalsystems3. Definition/identification of ‗challenge problem‘4. Design and construction of testbed, with physical modules distributed at multiple sites(simulations at all sites)5. Definition of interface specifications for modules6. Demonstration of module operations, validation of conformance with interface specification7. Validation of simulation model in normal and fault modes8. Demonstration of module interoperation in normal mode9. Demonstration of module interoperation in fault mode, and fault recovery using unified HMI10. Definition of reconfiguration scenario that was not originally envisioned11. Demonstration of reconfiguration through adding a new module and/or rearranging existingmodulesBROAD CHALLENGES AND BARRIERSTo realize the benefits of smart manufacturing, the exchange of information in manufacturing networkshas to be seamless, but such a comprehensive infrastructure is not yet in existence. 42 Some of the main39 Ibid.40 Ibid.41 Ibid.9
<strong>Cyber</strong> <strong>Physical</strong> <strong>Systems</strong> – <strong>Situation</strong> <strong>Analysis</strong>DRAFT – March 9, 2012challenges associated with implementing CPS include network integration, affordability, and theinteroperability of engineering systems. 43 Using a systems view of manufacturing is necessary to gain abetter understanding of what effects the changes in one element will have throughout the entire system. 44Smart manufacturing is only slowly incremented into portions of the manufacturing markets and does notprovide investment incentives. Most companies have a difficult time justifying risky, expensive, anduncertain investments for smart manufacturing across the company and factory level. 45 High costs areprohibitive to small and medium size companies, resulting in a slow adoption of smart manufacturingtechnologies in the supply chain that impacts the investment return for the companies that do invest. 46Pre-digital age control systems are infrequently replaced because they are still serviceable, and retrofittingexisting plants is more difficult than implementing CPS manufacturing technologies in new plants. 47There is a lack of industry standard approach to production management, and most companies customizetheir own software or use a manual approach. 48Today‘s production networks carry out all tasks related to the life cycle of the product, including design,engineering, fabrication, and maintenance functions and are, therefore, becoming increasingly complexand dynamic and require more sophisticated information integration. 49 This increasing complexity isbeginning to exceed the ability of both engineers and designers to fully control and optimize theirperformance. Traditional communication, control, and software theory cannot efficiently provide all thetools needed to analyze large-scale control networks. 50 For example, current network research oftenfocuses on connectivity and coverage issues assuming that network components are homogeneous; but inpractical terms CPS consists of both wireless and wired networks characterized with varying capacitiesand reliability. There is a need for a unifying theory of non-homogeneous control and communicationsystems. The heterogeneity of each device, in terms of memory, communication, and processing, shouldbe considered in the design of the CPS architecture to optimize real-time communication and reliability. 51Changes to the structure, organization, and culture of the manufacturing do not occur quickly, which issomething that must change if CPS technologies are to be successfully integrated. New smartmanufacturing technologies cannot be introduced incrementally or over a generation, but must be42 Smart Manufacturing Leadership Coalition, Implementing 21 st Century Smart Manufacturing: Workshop Summary Report (LosAngeles: Smart Manufacturing Leadership Coalition, 24 June 2011), https://smart-processmanufacturing.ucla.edu/about/news/Smart%20Manufacturing%206_24_11.pdf.43 Ibid; President‘s Council of Advisors on Science and Technology, Report to the President on Ensuring American Leadershipin Advanced Manufacturing (Washington, DC: Executive Office of the President, June.201),http://www.whitehouse.gov/sites/default/files/microsites/ostp/pcast-advanced-manufacturing-june2011.pdf.44 Robert E. Mansfield Jr. ., ―Extreme Manufacturing: The Future Manufacturing Enterprise‖ (Gaithersburg, MD: NationalInstitute of Standards and Technology, December 2010), http://www.nist.gov/el/upload/Extreme-Manufacturing_NISTWorkshop_Jan2011.pdf.45 Jim Davis and Smart Manufacturing Leadership Coalition, ―A Smart Manufacturing Public-Private Partnership Program‖ (LosAngeles: University of California – Los Angeles, January 2011),http://www.nist.gov/el/upload/SMLCtwopageprogramstatement1-0.pdf.46 Ibid.47 Ibid.48 Ibid.49 Simon. Frechette, ―<strong>Systems</strong> Integration for Manufacturing and Construction Applications‖ (Gaithersburg, MD: NationalInstitute of Standards and Technology, September 2011), http://www.nist.gov/el/msid/infotest/upload/SIMCAprogram2012.pdf.50 Yunbo Wang, Mehmet C. Vuran, and Steve Goddard, <strong>Cyber</strong>-physical <strong>Systems</strong> in Industrial Process Control (Lincoln, NE:University of Nebraska-Lincoln, 2008), http://sigbed.seas.upenn.edu/archives/2008-01/Wang.pdf.51 Ibid.10