An Industry-Driven Approach
By making the Smart Manufacturing Platform available to all potential users, costs typically required to develop business case specific control and management solutions will be significantly reduced. Smart Manufacturing Platform will be built in parallel with the development of SM System test beds for diverse applications. This approach will inform design concepts for Smart Manufacturing Platform development and create a push/pull commercialization strategy that accelerates its use in early adopter applications with significant measurable benefits. Test beds mitigate risks that individual companies would face in researching and implementing smart manufacturing on their own. SMLC brings together industry, government, manufacturing and university resources to accelerate the development of smart manufacturing approaches and technologies.
A prototype of the Smart Manufacturing Platform is being developed under the DOE Project. Additional test beds leverage the outcomes from the grant.
Test Bed Examples
In all of the illustrated Smart Manufacturing test beds, success with the first application provides a set of contextualized data defined for a particular objective, deployment experience with data management, analytics and modeling, actionable use of the results relative to a metric, and standards. This foundation establishes a basis for expansion with much greater expectation of success. Additionally, success with a location can provide a template to scale the IT infrastructure to additional plant sites, which can lower the cost and risk of development and implementation. As a result, test beds together with SM Platform technology can mitigate short and long term investment risks, improve paybacks, provide methodologies for replication, and shorten development times for the adoption of Smart Manufacturing systems.
Steam Methane Reforming
Continuous flow process, energy intensive, furnace operation – Steam Methane Reforming (SMR) is an energy intensive process that uses catalytic reactions in steel tubes in large scale furnaces heated to about 1,000 °C to produce hydrogen, synthesis gas, ammonia, and methanol in over 900 facilities worldwide. In the test bed, furnace operation is managed using Reduced Order Models (ROMs) with high fidelity modeling to spot-check and periodically validate the ROM. Continuous collection of furnace temperature data with infrared cameras could allow individual burners to be fine-tuned for dynamic optimization of the heat distribution. However, high fidelity modeling for production involves different time requirements since the model cannot be solved within the response time associated with the process and control system. An SM System extension of the control system can provide real-time updates to a ROM using windows of operation validated with the high fidelity CFD model running multiple parallel predictive sessions on a regular basis. A 20% reduction in wasted energy and corresponding improvements in productivity, fuel usage, and GHG emissions have been projected. The SM Platform substantially reduces risk, brings payback within thresholds and offers the methodology to replicate. The SM Platform also shortens the model and data management configuration process to bring the SM system into production sooner.
Fabrication of Precision Metal Parts
Discrete, batch process, energy intensive, fabrication operations – Fabrication of precision metal parts involves a series of forging, heat treatment, and machining steps that convert raw material into parts with customized geometrical and metallurgical specifications. The heat forging and heat treatment processes are energy intensive and frequently operate furnaces in excess of 750 °C. This manufacturer produces numerous products with multiple process changeovers and varying orders from customers. New furnace control systems and optimization of CNC machining have already resulted in significant energy savings and productivity improvements. Beyond these, the SM System will use in-situ measurement and/or inference analysis to model the effects of heat treatment on metallurgical properties together with operations. The data will be used to enhance and dynamically manage the metallurgical structure of parts and energy together to improve downstream machining productivity, reduce rejects, and dynamically manage fuel and power use in all operations. There is considerable untapped opportunity for improvement in economics, gas and electricity usage, and machine maintenance across the entire product line. The SM Platform makes in production HP model-based solutions an operational, business and financial reality and is itself used to build an initial model of the line operation that can be readily grown in sophistication and capability… IT also facilitates trialing different software solutions within an assembly of models and analytics solutions and developing respective model and data configurations.
Food and Supply Chain
Cross company food supply chain interoperability – Input product qualification from supplier to buyer, as well as traceability of product from buyer to supplier, are important manufacturing practices required by federal and state governments in the food industry. Today the steps to ensure a lot of grain meets or exceeds regulatory requirements, company requirements, and quality standards still rely heavily on a paper-based certificate of authentication (CoA) from the supplier verification testing by the buyer. The SM System in this test bed integrates several key steps in the overall purchasing, shipping, and receiving process to: 1) allow a buyer to recast the CoA for supplier product into data, 2) manage shared data in accordance with agreements between the supplier and buyer; 3) map the CoA data from multiple suppliers into the units and definitions required by multiple buyers; 4) interface securely so the buyer receives the supplier’s data in time to incorporate variability into manufacturing readiness; 5) facilitate continuous improvement of operations between the supplier and buyer. Significant cost benefits can be derived from the use of raw material data to adjust production processes in advance of material delivery, i.e. reduced inventory. The SM Platform provides initial capability for managing selected data with appropriate security, privacy and policy from multiple companies. As an electronic CoA grows in sophistication with more sensor data, the SM Platform makes it possible to assemble analytics and modeling for cross supplier and buyers needs and for cross company energy, environment, transportation, opportunities and management of regulatory requirements.
Process and Fabrication
Configurable data and models for rapid analytics, model development, and approaching “Big Data” – This process and fabrication test bed focuses on the modeling processes involved in sensing, analytics, and platforms, and their integration. The specific goal is to reduce the time and effort for making data oriented enterprise decisions, 90% of which are devoted to gathering the right data, contextualizing it, and setting up analyses sources of diverse data such as plant historians, real-time process state and part quality data, equipment specifications, and supply chain databases. The test bed is using the SM Platform for progressive development and application of MI by growing data and modeling with increasingly sophisticated, but well defined performance objectives, to always drive contextualized data and modeling needs. An analysis of Smart Manufacturing Platform economics (compared to no platform) shows that cost based risks and actual costs of developing an SM data analytics application are reduced as much as 25% for the first model of staged modeling approach. Replication and reuse of Smart Manufacturing applications for similar operations could be reduced by as much as 60% for the first replication.
Additional Test Beds
If you are interested in participating on a project. Please contact Swinkdenise@aol.com for more information.