Streaming Data and Event Infrastructure - OSIsoftcdn.osisoft.com/corp/en/docs/whitepapers/WP... · Streaming Data and Event Infrastructure Last Saved May 4, 2012 2 of 20 ... However,

Streaming Data and Event Infrastructure

Last Saved May 4, 2012 1 of 20

OSIsoft, LLC. 777 Davis Street, Suite 250, San Leandro, CA 94577

Unpublished -- rights reserved under the copyright laws of the United States.


Version 2012.10

Table of Contents

1 Introduction ......................................................................................................................... 2

2 Infrastructures ..................................................................................................................... 2

3 The Power of Real-time Data .............................................................................................. 3

4 Business Challenges that need Data ................................................................................. 6

5 Customer Success Stories ................................................................................................. 9

6 Ingredients of an Streaming Data and Event Infrastructure ........................................... 15

7 Summary ............................................................................................................................ 16

8 References ......................................................................................................................... 18



1 Introduction

Companies today are in need of a “Streaming Data and Event Infrastructure” to enable their business to respond quickly and decisively to a wide range of issues. The maxim ‘what gets measured gets done’ still applies to any organization that intends to drive its operations to a higher level. However, building a responsive data driven business during times of great change and uncertainty will require an updated perspective, new tools and skill sets to make access to data part of the organizational culture. Companies must also empower employees to take action on the data. This whitepaper describes a “Streaming Data and Event Infrastructure” required by business today in order to improve performance. This white paper will first look at what we mean by infrastructure and how it has been a key enabler to business. The “Power of Data” to achieve situational awareness, empowers continuous improvement and provides for sustainability and innovation. Today’s key business challenges and how data is used to address them will be discussed. Customer examples of using an infrastructure approach to manage real- time data and the resultant success will be given. Lastly the components of a “Streaming Data and Event Infrastructure” are highlighted.

2 Infrastructures

Infrastructures have a long history of making great change in business and society. The Railroad was an infrastructure that opened the West, allowed cities to grow up along it or fade away if they did not or could not leverage. In the 1950’s President Eisenhower supported the Interstate highway system that made our society more mobile. The municipal water systems gave rise to more modern and cleaner cities, the power grid enabled electronic society of appliances for all aspects of our lives. The Internet has connected the globe, changed the face of commerce, and brought information to the masses. Today’s wireless networks have truly mobilized our workforce giving instant access to data, information and personal resources. OSIsoft believes in the power of an infrastructure as enabling technologies. Their principles have invested in partnering with the city of San Leandro where it is headquartered to provide access to high speed, low cost 10Gbps internet as an enabling technology for business development in the city for the Lit San Leandro project. A single application or requirement need could not have justified any of these great infrastructure projects however the results have been paid back many times over their investment. However, we often see businesses approach the access to operation data needs as a contained, singular solution that must be justified on isolated business requirements. This results in point solutions that often isolate data, limit access and fail to meet changing business requirements. Today’s businesses are required to operate fast and globally, often making decision on limited data, e.g which plant has capacity, where is inventory, which assets have low reliability, which

http://www.sanleandro.org/depts/cd/bizdev/broadband/lit_san_leandro.asp



products are costly to make, and the list goes on. How much better would these decisions be if the people had all the data, everywhere and anytime? Often we do not know what we need until the problem arises and then it is too late to question if we collect the data. Our world today is instrumented all over. Sensors were once applied to only the most costly equipment however today we are now able to obtain measurements from a large range of devices and systems economically. OSIsoft proposes and has since 1983 that a business today should collect all the data all the time at the original data fidelity because aggregation and averages can mask behavior or transient upsets. Streaming data systems are able to collect and store this data without excessive costs within a real-time infrastructure for data streams and events. The building of such an infrastructure can provide companies a competitive advantage when they have data required to focus on innovation and sustainability while maintaining situational awareness and continuous improvement.

3 The Power of Real-time Data

“Power of Data” which is a topic that is universal, transcends all frontiers, transcends all industries, transcend all activities and is making it in our life, in a big way, called “Big Data”. The topic of data is now ubiquitous, a daily conversation in all spheres of business, technology and society. The etymology of data, the verb “dare” in latin, refers to what we “are given” to analyze, understand, and solve problems. A recent New York Times article “Age of Big Data” proposes that we are in “the age of big data.” Big data, says the article is “shorthand for advancing trends in technology that open the door to a new approach to understanding the world and making decisions.” The article quotes IDC, a technology research firm as estimating that data, by volume, is growing at 50 percent each year. There are now countless digital sensors worldwide in industrial equipment, automobiles, electrical meters and shipping crates. They can measure and communicate location, movement, vibration, temperature, humidity, even chemical changes in the air according to the article. Communicating sensors with computer intelligence gives rise to improved information access so that decisions will increasingly be based on data and analysis rather than on experience and intuition. In the last decade we often used catastrophic event metaphors such as Data Eruption, Data flood, Data Tsunami, Data deluge, to describe how overwhelming data had become. Some have referred to Data Overload, Data Obesity, Noise, Distraction or the need to go on an information diet to describe the challenges of dealing with large quantities of data. As the information age, described as “the third wave” by the futurist Alvin Toffler, settled in and computing and networking were commoditized, we started to realize that data, lots of data could be turned into an opportunity and started using the term “Big Data” to describe data quantity that exceeds the processing capacity of conventional database systems.



OSIsoft customers and PI System users, have been dealing with this data reality for many years. They know that “data” is an essential element of better decision making, the foundation of operational excellence, innovation and sustainability and is essential for maintaining situational awareness and readiness. We estimate that PI System users, taken collectively manage over 235 MM data streams at this moment and customers have been historizing their PI System data for over 20 years. The number of data streams has exploded in recent years and will continue to grow exponentially driven by the internet of things. We estimate at 300 MM the number of active data streams that the PI Systems will be handling by 2012 year end. Timely, easy and enterprise wide access to streaming data and events is a requirement of the modern enterprise and a very tangible manifestation of the “Big Data” phenomenon which is receiving so much attention lately. We are moving into a data, measurement and insight based world where experience, knowledge and data are combined to drive value creation, drive latency out of decision making understand change and increase predictability of targeted outcomes. Today, entire businesses are built around data and these businesses thrive and support economic and social development. One of the best examples we experience every day is the management of the electrical grid which is largely dependent on streaming data and events, in very large quantity, to achieve the availably and the reliability of electrical energy supply. That data that supports these businesses has to be as reliable and available as the service they provide, in this case electricity. This is one of the reasons the industry is hiring the PI System and OSIsoft to do this job. Large data sets like the ones found in PI systems have notable characteristics:

VOLUME - The volume of data is large – a function of how many sensors, their sampling rate and how long data has been collecting;

VELOCITY – it varies greatly, can be very high and will be getting higher; data may arrive synchronously or asynchronously; latency can be negligible or significant and important.

DIVERSITY - Wide diversity of data types come into play - analog input, digital Input, conventional and intelligent sensors, external data sources like weather, prices, non persisting business data, etc.

STRUCTURE - Data payloads from intelligent sensors carry the data and information about the data: its structure, its lineage, and its state. Systems must recognize and leverage this ‘meta’ data creating and updating structure on the go.

ANALYTICS - Pre and post processing driving real time analytics are the norm and used on a large scale at high speed and are an essential part of the maintaining deep situational awareness.

What is the power of all that data?

This power of data is knowledge preservation.

This power of data is to enable assets to speak to us?



The power of data allows important signals to manifest themselves through trends and analysis rules;

This power of data drives predictability;

This power of data helps figuring out cause to effect relationships as in condition based maintenance, RCM;

This power of data is to measures outcomes or impacts of our decisions;

The power of data augments our ability to decide, resolve issues, innovate and gain insight.

The Power of Data is universal, available to everyone, augmented by collaboration, and, is moving us into an era of rapid innovation and progress. So, is Big Data hype or reality? Real Big Data requires new skills and new roles such as data analysts, data scientists, data experts, etc. But, we see that the interest for these new roles is still lagging the “Big Data” phenomenon. Fortunately, OSIsoft customers have been data driven for many years and have a head start in the practice of the data analysis or data science. What is this head start worth? Dr. Brynjolfsson and his colleagues at MIT reported in one of the first research works conducted on data driven decisions that this technology can boost productivity and output by 5 to 6 % and drive the adoption of information technologies enterprise wide. With the power of data at hand, we are entering an era where data driven decision making will become not only a common skill but a skill critical for success, both for people and organizations. Situational Awareness The availability of data allows a business to maintain situational awareness, to know where your business is and the external forces that affect it. Whether a business is looking for its production numbers, quality, emission, productivity, resource consumption or commodity pricing and market fluctuations or customer requirements the data is available. An exact picture of where you stand and comparisons to similar situations is available for analysis and insight. With this much data, obtaining it by exception via notifications as events occur has become critical. The data is available for multiuser access and collaboration for better solution resolution. Detail analyses for planning and what if studies are now supported with actual data. The data is easily visualized and placed in context lowering the cost of curiosity and managing the risks and opportunities of the business. Continuous Improvement Continuous improvement requires knowledge of the current and past situations in order to identify the root cause of the problem. The availability of current data as well as historical data



provides the context to identify problems and propose solutions for continuous improvement. One must also have the ability to implement six sigma procedures of measure, analyze, control and optimize. Visualization tools that provide methods to performance benchmark, display KPIs and dashboards are key to measuring improvement and making data available to everyone to fuel a culture of improvement. Tools for visualization also allow the user to discover new patterns, insights and are the fuel for innovation. The socialization of data allows for collaboration and problem solving. Data must be provided in context with easy tools to lower the cost of curiosity and cross specialty data access. Foundation for Sustainability and Innovation From its “green team” origins, corporate sustainability has found its way into the executive suite and corporate boardroom. Many organizations now have a chief sustainability officer, and some of these same companies routinely issue reports on their progress toward sustainability. However, largely missing from the sustainability calculus is a way to move into the mainstream of day-to-day workplace decisions. For sustainability to reach the next level, operations staff who make resource decisions need feedback and visibility on the results of their actions, and senior executives need a way to gauge performance so they can reward improvements and troubleshoot lingering problems. Sustainability is a platform for managing business in an efficient and holistic manner. It provides resource and production decision-makers with tools and support to make better decisions. Key benefits of a focus on sustainability are resource, water and energy efficiency, pollutions abatement and asset optimization. All of these are empowered with the access to data. Today sustainability data is collected and available with large time delays. Tomorrow we have the potential of controlling our business to sustainability goals with real-time data allow actions to take place to improve situations rather than just report. Records of sustainability data will take on an importance similar to financial data. The best innovation comes when people ask what if, why not, how we can in a process that encourages action. All of these questions are helped with the access to data that allows for the modeling, scenario testing and results validation. Innovation is inspired by seeing patterns in data and studying relationships between data. When the access to data is lowered and data is available across disciplines new ideas can be proposed, investigated and proven.

4 Business Challenges that need Data

Today’s business faces common challenges no matter what the industry is. We reviewed 73 annual reports from companies in Pulp & Paper, Oil and Gas, Utilities, Pharmaceuticals and Life Sciences, Metals and Mining and Chemicals and Petrochemicals compiling what are identified as business challenges. This list shows the clear advantage of infrastructure over a single project based solution due to the breadth of the issues.



Cost Management Business must control their capital and variable costs. This means how the most can be achieved from available assets and prolonging the lifetime of equipment and facilities. What facilities are operating better and which ones have room for improvement and where and when to invest in new facilities and equipment. All of these require access to data and operational statistics. There are also the large variable costs of energy and water and how the process of using both can be controlled and optimized. The use of raw materials and the recycle and repurpose of waste material for financial gain require process optimization and innovative solutions. Businesses continue to look to acquire and divest as required by the changing marketplace and global needs of operations. The data associated with these facilities must be easily integrated into the corporation or separated. This requires a flexible data infrastructure that easily integrates with existing systems. Sustainability Today’s corporation’s actions are more visible and they are being held to a new standard of corporate citizenship by customers and governments. This means compliance with regulations for health, safety and the environment. Data supporting the corporate result is required to become more visible and timely. In some industries the carbon footprint must be associated with finished products and customers are making buying decisions based on the data. Sustainability requires operations to save energy, optimize raw material resources and manage water use while minimizing waste production. Communication of corporate status on sustainability is becoming a more frequent activity and needs will be changing to operate for sustainability rather than just reporting after the fact. To move to sustainable operations data will be required in real-time to allow for actions to maintain sustainability objectives. Manufacturing and Operations Manufacturing is required to produce a high quality product at the lowest cost while complying with environment and legal requirements. There is the need to plan production across the whole manufacturing space while operating an individual facility at its optimum conditions. This requires accurate, timely information to be shared across facilities while optimizing and responding to operational demands within a facility. Data must be readily available and shared. Assets must be controlled, maintained and serviced to meet the demands of production while assuring the lifetime of the asset. There is the need to cost effectively increase production will maintaining operational integrity. Often new demands in flexible manufacturing must be met to respond to changing product and global market demands.



To meet these demands rigorous benchmarking, upgrading and debottlenecking of facilities is required, all of which require operational data for engineering studies. The true cost of production for products must be known. Enhanced production, operation and decision making for continuous improvement require access to data. Performance improvements and development is made with continuous feedback, attention to global standards in operation and implementing state of the art manufacturing for quicker response to market demand and reduction of operating cycle times. Data availability in real-time is key to all of these initiatives. Globalization Companies are global today. This brings the need to manage the business globally while responding to local requirements, laws and customs. Global business puts more pressure on supply chains, inventory management while meeting trading and market conditions. Balancing a global asset base, inventory, customer demand, regulatory and economic environments require access to data across the entire company. Often decisions need to be made rapidly. The access to data allows for these decisions to me be made from information rather than relying entirely on experience and intuition. Workforce The workforce must be empowered with the right data at the right place in the right time to make decisions. They must have the ability to work across organizations for collaboration and the access expert specialists. Today’s workforce is highly mobile and always on call. They need to have the data as well as the analysis tools available on demand and from personal devices. Technology Technology and the need for data analysis are moving organizations towards the convergence of Information Technology and Operation Technology. These groups previously worked as silos within an organization. Today both groups are learning from each other to meet the demands of the business. Trends and requirements in Big Data, Analytics, Security and Reliability are being met with technology. Corporations are seeking technologies to help their people work smarter while maintaining the security of corporate IP. Innovation Whether it is product or process innovation companies are seeking the next breakthrough that will give them new opportunities in the market. Often data plays a role in being the catalyst to provide insight into existing conditions that open pathways to new opportunities. The flexibility to look at data in many ways and drill down into details of data at their original fidelity are some of the requirements when it comes to looking for innovation in process and product



improvements. According to Michael Porter of Harvard Business School and Director, Institute for Strategy and Competiveness, the challenge for competiveness has shifted from the last decade of restructure, lower cost and need to raise quality to the innovation challenge. He says advantage must come from the ability to create and then commercialize new products and processes. A streaming data and event infrastructure is a critical requirement to address all of these business challenges.

5 Customer Success Stories

Alcoa

Alcoa is the world leader in the production and management of primary aluminum, fabricated aluminum and alumina. Alcoa’s sales in 2008 were $26.9 billion. They have 63,000 employees in 31 countries and have been named one of the most sustainable corporations in the world. Alcoa also makes a very sustainable product: more than 70% of the aluminum ever produced is still in use. Alcoa is in the very competitive global market of producing aluminum. Energy is 40% of their production costs and they run an 800MW generation plant to power their Warwick smelter. Two years ago, the Warwick plant decided to tie into their electricity provider’s demand response system to find new ways to manage their energy use. Using the PI System, Alcoa balances between consuming power and selling it to the electric utilities depending on which is more profitable. “Demand response is huge for us,” says Brian Helms of Alcoa. In the first year with the help of their PI System, they made 1,800 small changes in generation to earn millions of dollars in new revenue that would otherwise not have been possible. As Brian Helms, Alcoa project manager, put it, “Incremental changes, that’s what we found the PI System is doing for us. These little things that we can add to what we are already doing, to put the information out to the people making the decisions, and we reap the benefit of those decisions. That’s the power of the PI System.” Alcoa has the PI System installed in 20 facilities in 8 countries utilizing 400,000 tags, 27 PI System servers, 73 interfaces and 600 clients. Most recently the PI System was installed at Alcoa Aluminum’s largest North American smelter located in Newburgh, Indiana, USA. Adjacent to the smelter is Alcoa Power Generation’s Warrick Power Plant who’s 732 MW’s feeds the smelter and rolling mill processes. Prior to participating in the Midwest Independent System Operator’s (Midwest ISO) Energy Market, selling excess MW’s or buying needed electricity took resources to negotiate and manage those buy and sell transactions. Seeing an opportunity to participate in the upcoming Midwest ISO Ancillary Services Market (ASM), Alcoa Power Generation purchased the PI System as their underlying data infrastructure. When the ASM went live in Alcoa Power Gen became the only type II participant in the form of Demand Response. Alcoa allows Midwest ISO to send signals that are received



by Alcoa’s Smelter Control System to either increase or decrease load. An increase load signal means that the smelter is consuming excess MW’s from Midwest ISO’s system which also results in an increase in aluminum production. A load decrease signal results in a drop in aluminum production thereby diverting some of the Warrick Power Plant’s electricity into the Midwest ISO system to make up load demand of the grid. Either way, Alcoa wins. With common data, Alcoa talent across the globe will engage in collective innovation and the pursuit and sharing of best practices. In order to focus Alcoa Talent on management of the plant, measurement data must meet the following criteria: Accurate, Real-time, Correlated and Historized. Sharing a common language with other plants will allow Alcoa to truly realize the Alcoa Advantage. A discovery in one plant can result in procedure, training, and policy changes in the other plants. SMART measures will become common measures. They will allow Alcoa to share best practices from one plant to another.

IBM

IBM Burlington is a large semiconductor manufacturing site which consumes 3.2 million gallons of water PER DAY and 446 million kilowatt hours of electricity annually. The sites potable water is supplied from the Champlain Water District. This potable water must be purified to meet the strict Ultra Pure Water (UPW) quality requirements needed for semiconductor manufacturing. UPW is 10 million times cleaner than the raw potable water. By using the PI System, IBM has achieved over $10M in annual savings, reduced water usage by 27%, reduced operating costs, and minimizing environmental impact while increasing manufacturing capability by 30%. The IBM Corporation is using Advanced Water Management techniques to drive cost improvements required to compete in a global market. These techniques are applied to Ultra Pure Water systems, heating and cooling waters, potable, non-potable and waste water treatment in semiconductor manufacturing. Water systems are closely monitored and data is collected and analyzed to support the decisions necessary to achieve IBM’s desired water management goals. These goals include strict control of water quality, reduction in water consumption, reduction in energy consumption and overall lower cost of operation. The IBM plant has achieved a 27% reduction in water usage, $3M in annual savings while manufacturing output increased by more than 30%. IBM’s data driven water management techniques have wider application that can be applied to address water issues for water stewards, distributors, industry and consumers. Water management challenges are largely a data management challenges. From the water resource all the way through to the water consumer, to the wastewater treatment operations and back to the resource there exists vast amounts of water data: usage, sustainability, quality, pricing etc. Organizing the data, analyzing the data and presenting the results in easily accessible formats transforms the data into manageable pieces of usable information. This information is the basis for sound decision making around water issues.



The challenges are extracting the information in an efficient and meaningful way, presenting usable information to the proper levels of the organization and making data supported decisions to move towards established goals. Although the water industry is data intensive, to a very large extent data management systems have been lagging behind some more advanced industries such as petrochemical and semiconductor manufacturing. As water management issues become more critical the application of these techniques become appropriate. IBM uses the OSIsoft PI System for data collection, organization, analysis and presentation of data. IBM’s advanced water management data analysis tools were developed borrowing techniques from IBM’s manufacturing operation. These techniques allow for the transformation of vast amounts of raw water data into manageable amounts of useful information. These advanced water management techniques can be successfully applied to water management issues for ecosystems, utilities, industries and municipalities. KPIs provide the primary information tool to drive an organization towards its water management

goals. KPIs were developed to be user appropriate at various levels of the organization. The IBM

Burlington site has reduced over 2000 water monitoring data point to 80 KPIs. These 80 KPIs are

analyzed automatically, abnormal circumstances are brought to the attention of operations and drill

down capabilities within the software allows for quick troubleshooting of excursions.

IBM’s water treatment goals can be simply described as continuous operations, maintaining water quality and operating at low cost. There are over 400 million data packets per day being analyzed. This amount of data would be overwhelming without advanced techniques to transform it into usable information. IBM has applied OSIsoft data collection and visualization along with KPIs, SPC and Cost Modeling to support IBM’s water management goals. These techniques are used in combination with more conventional real time data trending and correlation. The goal of these techniques is to transform the 400 million packets of data into a manageable amount of useful information. Through the use of computer driven data analysis only important information is extracted and brought to the attention water system engineers, technicians and operators. There has been a 27% reduction in water purchases and associated purchased water cost. Since much of the water saved was Ultra Pure Water there are additional savings associated with the treatment steps. There is a strong relationship between water and energy. This was important for IBM as it provided additional justification to implement projects. There is capital cost avoidance related to supporting the increased semiconductor manufacturing output. The efficiencies gained allowed support of increased manufacturing with existing infrastructure.

Sappi Fine Paper

The pulp and paper industry is incredibly competitive—retention and expansion of market share are key factors to success. Not only is this industry capital-intensive, it also requires the balanced management of multiple resources. Add to that the recent interest customers have in knowing more about where their paper comes from, and the business need for rich data is



clear. Sappi Fine Paper is the preeminent producer of coated fine paper in North America — with a production capacity of 1.3 million tons of paper annually and marquee customers ranging from some of the world’s leading fashion magazines to premium catalogs. The company has long made sustainability a key component of its business strategy, increasing profits by improving operational efficiencies and reducing waste, and winning new customers by producing some of the industry’s most environmentally preferable papers. Sappi North America has been using the PI System since 1984. The company currently has 150,000 to 160,000 active PI System tags, the majority of which come from shop floor devices in its mill facilities.

“Operational staff look at this data and make minute-by-minute decisions based on what

they’re seeing,” said Thomas E. Bolen, Sappi Global COE Manager for the PI System.

“I use the PI System every day to make sure the demands of the mill are met—whether it be electricity, steam, air, or water,” said Todd Anderson, utilities operations manager for Sappi North America. In addition to supporting Sappi’s day-to-day operations, access to real-time data through the PI System has delivered measurable cost reductions and improvements to environmental performance. “When you have these sorts of capital assets in place, you have to get the most out of them in terms of efficiency,” said Laura Thompson, Director of Sustainable Development and Technical Marketing. “Access to data helps us understand how we’re using resources.” Sappi staff in various departments use the PI System daily to be more conscious of environmental impacts and operational costs, and to make real-time decisions that often reduce both. In the case of power consumption, for example, the PI System enables operational staff to determine whether to buy more energy from the grid or to ramp up steam production in its own boiler. Using data from the PI System, Sappi continuously balances the power it generates with the power it purchases which often keeps both costs and emissions low. This not only delivers bottom-line savings, it has also helped the company deliver on its goals to use more renewable energy.

Cargill

Cargill is an international provider of food, agricultural, and risk management products and services. The company was founded in 1865 and had revenues of $71 billion in 2005. With 124,000 employees in 59 countries, the company is committed to using its knowledge and experience to collaborate with customers to help them succeed. They put in their first PI System in 1996 and now have 66 in the US and Canada, 11 in Latin America, 51 in Europe and 9 in Asia and Australia as of 2012.



Michael Smith, energy engineer, Grain & Oilseed Supply Chain North America at Cargill, Inc. Presented at OSIsoft UC, “Using data to set operational goals and incentives to change and reward behavior to achieve business objectives.” Cargill discussed the importance they place on the strategic use of reliable operations data to set effective performance goals and employee incentives to drive behavioral change. Cargill is looking to reliable data to drive their organizational targets for sustainability. They have outlined corporate goals for 2015 which include a 5% improvement in energy efficiency, growing their use of renewable to 12.5%, a 5% improvement in GHG emissions and a 5% improvement in fresh water efficiency. These goals are set through detailed engineering studies that evaluate their processes. The Engineering models address mass and energy balances and explore new engineering concepts to attain plant performance improvement targets. Once the engineering has been done the models become the targets for operation. The PI System is used to monitor performance, explore conditions and perform root cause analysis of the gaps. The plants then develop solutions for improving such gaps and document the reasons and exceptions. Reliable data has been a method for Cargill to set their targets and drive the organization towards those targets.

PJM

Founded in 1927, PJM Interconnection is an independent, federally regulated organization that operates from Valley Forge, Pennsylvania, USA. Three original members were utilities from Pennsylvania (P), New Jersey (J), and Maryland (M) – thus PJM. They coordinate and direct the operation of the region’s transmission grid and administer a competitive wholesale electricity market. PJM is a Regional Transmission Organization (RTO) that manages bulk power system reliability with PJM is a Regional Transmission Organization (RTO), coordinates the movement of wholesale electricity in all or parts of 13 states and D.C, operates wholesale electricity market and manages the high-voltage electricity grid to ensure reliability for more than 58 million people. PJM operates with Peak Demand of ~158,000 MW, Peak Capacity 0f ~180,000 MW, a territory that includes 6,000 substations with 61,000 miles of transmission lines (69KV-765KV) and over 1,200 generators. Currently, PJM uses 450,000 PI System tags across about 70 computer servers. One interesting application is called Interconnected Reliability Operations Limits. It displays flows and contingency flows on critical transmission equipment in tabular and graphical form in PI ProcessBook with timers for compliance purposes and drill-down displays to isolate causes of excess flows. The application also contains approved operations procedures for controlling excess flows.

Sempra Energy

“Sempra Energy is ranked the most intelligent utility in US and most reliable utility west of the Mississippi,” says Patrick Lee, Vice President of Energy Supply. Sempra Energy’s California utilities, San Diego Gas & Electric Co. and Southern California Gas Co., serve more than 20 million consumers. And its other businesses – Sempra U.S. Gas & Power and Sempra



International – develop and operate critical energy infrastructure and provide gas and electricity services in North America and South America. Sempra Energy required real-time visibility to data across the whole value chain. They wanted to increase reliability and situational awareness of all operations across the utility. The implementation of the OSIsoft PI System allowed for a single infrastructure to address these requirements for generation, transmission, distribution, LNG storage, renewable and pipelines. Each business area developed applications specific to the business including monitoring, situational awareness, alarms and notifications. This has resulted in benefits of increased reliability, data for planning and forecasting, environmental reporting, asset maintenance and overall grid operations.

Johnson & Johnson

Johnson & Johnson is an American multinational pharmaceutical, medical device and consumer packaged goods manufacturer founded in 1886. The OSIsoft PI System became an integral part in the integrated manufacturing system that Johnson & Johnson was implementing. Their objective was to provide a manufacturing automation system that would eliminated the need for manual data collection. They wanted to eliminate transcription errors and time consuming cross checking of data. There was need to consolidate the building alarm reporting and management with process, laboratory, warehouse and utility systems. This would minimize the data gather time as well as the time required to perform investigations. Consolidation of data visualization was required for improved process monitoring and historical batch analysis. This was an enabling step for integrated manufacturing design. The OSIsoft PI System provided interfaces to batch and manufacturing execution systems. The PI System allowed for the compete modeling of the facilities and batch process maintaining the data in operational context. The resulting system delivers improved efficiency of production and administration, improved yield and reduced cost. The overall system gives better situational awareness of the manufacturing process, provides the data for continuous improvement while helping to control costs of manufacturing and providing the data and reporting requirements for the FDA.

Bristol-Myers Squibb

Bristol-Myers Squibb is global Bio-pharma company whose mission is to discover develop and deliver innovative medicines that help patients prevail against serious diseases. There need to improve operations included making the operators jobs easier to allow them to focus on the batch process with seamless connectivity behind the scenes to the appropriate



systems. They needed to monitor all manufacturing activities and perform batch reporting by exception. The OSIsoft PI System became the backbone of an integrated solution that redefines the manufacturing operation philosophy. The resulting systems reduced overall cycle time, provided better coordinated information flow between the back and analytical activities. Direct communication between different automation systems was facilitated. There was full capture of the batch recipe for documentation. The system has resulted in more reliable and repeatable batch execution for FDA compliance.

Chevron Corporation

Chevron is one of the world’s largest integrated energy companies. It is headquarter in San Ramon, California and conducts business in more than 100 countries. Chevron is engaged in every aspect of the oil and natural gas industry, including exploration and production, manufacturing, marketing and transportation, chemicals manufacturing and sales, geothermal and power generation. Their business requires robust access to real-time process data across all of these business units that can be aggregated and shared in a centralized repository. This is a large volume of streaming data points and events that must be captured in real-time. The data required a metadata layer to put the information into context of geographical location and business units as well as assets and operations. The access to this information has resulted in reduced turnaround time for projects, rationalized standards across facilities and business units and has enabled centralization of development and deployment across the corporation.

6 Ingredients of an Streaming Data and Event Infrastructure

At a functional level, a data delivery infrastructure should connect end-users that can make operational decisions with access to real-time key performance indicator (KPI) data for their areas of responsibility. It should also provide executives with dashboard oversight of overall performance. Both line employees and executives should have the ability to drill down into detailed views of individual devices and equipment that they control. Key capabilities for a data infrastructure include: 1) data collection, 2) data management, 3) analytical support, 4) data delivery, and 5) support for intuitive visualizations. Each of these capabilities will now be discussed in more detail.

1. Data collection interfaces make it possible to gather data from sensors and industrial process control devices in real time. Sensors can include purpose-built meters that



measure the consumption of energy, fuel and other resources; sensors that monitor exhaust such as heat waste, carbon monoxide, SOX or NOX; and industrial process control devices and systems that measure and control production processes. Support for a wide range of devices makes it possible to source data as needed to build sustainability metrics and dashboards and to support value-added analytics.

2. Data management capabilities are needed to gather and consolidate different data streams from devices so the data can be merged to gain a common view. Data management must also include the ability to create relationships between devices, processes, and people. This stages the data for more intuitive access and use.

3. Analytical support should include the ability to aggregate real-time and historical data and events into intuitive formats, including key performance indicators for different sustainability objectives such as conserving resources, reducing pollution, maximizing yields and ensuring maximum asset performance and longevity.

4. Once data is consolidated and organized, it is ready for delivery (push) and shared access (pull) by the people and processes that can benefit from it. Data delivery capabilities should include support for real-time feedback to personnel managing processes and systems that have a direct impact on objectives, and for delivering summary-level data to senior managers and executives interested in a more consolidated, birds-eye overview.

5. Examples of intuitive visualizations include process control dashboards that show key values in a format that resembles the dials in the cockpit of an airplane, graphs and charts that show trends and anomalies, and standard reports that can be further customized to meet specific needs. The ability to download and manipulate data in comfortable and familiar desktop tools like Excel and more powerful statistical analysis tools is also important for empowering end-users to conduct their own analyses, discover and innovate from data.

A comprehensive data infrastructure provides these end-to-end capabilities in a seamless and easy-to-use fashion, but with depth of functionality and the ability to tune and configure each capability to meet specific needs. The infrastructure must provide a secure data foundation that is reliable, highly available, scalable while being extensible and able to interoperate with other systems including enterprise systems.

7 Summary

With continuous access to KPIs and real-time performance data, it is possible for workers and executive management to incorporate performance as a key element in the work that they do. Baselines that provide data on historical performance are an excellent starting point. Ongoing performance can be compared to historical baselines, and the baselines can keep evolving to support continuous improvement.



Organizations that understand their operational footprint and set ambitious objectives – and then back those objectives up with key performance indicators and a robust “Streaming Data and Event Infrastructure” – will be well on their way to lasting gains in performance. The key is employee empowerment that enables bottom-up action and top-down visibility using real-time performance data. Visit http://www.osisoft.com/ to learn more about the OSIsoft PI System, a “Streaming Data and Event Infrastructure”.

http://www.osisoft.com/



8 References

1. Lit San Leandro (2012, March). Retrieved from

http://www.sanleandro.org/depts/cd/bizdev/broadband/lit_san_leandro.asp

2. Lohr, S. (2012, February). The Age of Big Data. The New York Times Sunday Review. Retieved

from http://www.nytimes.com/2012/02/12/sunday-review/big-datas-impact-in-the-

world.html?pagewanted=all

3. Byrnjolfsson, et al., MIT, How does Data-Driven Decisions making Affect Firm Performance.

2011. Retrieved from http://www.nytimes.com/2011/04/24/business/24unboxed.html

4. Stern, S. Porter, Porter, M., National Innovative Capacity, Paper Harvard Business School,

Retrieved from http://www.isc.hbs.edu/Innov_9211.pdf

5. Helms, B. (2010, April). Industrial Scale Demand Response. Paper contributed to the OSIsoft

User Conference, San Francisco, CA. Retrieved from http://www.osisoft.com/Templates/item-

abstract.aspx?id=5059

6. Chapman, J (2010, April) Advanced Industrial Water & Energy Management Saves $10 M

Annually: IBM Vermont Case Study. Paper contributed to the OSIsoft User Conference, San

Francisco, CA. Retrieved from http://www.osisoft.com/Templates/item-

abstract.aspx?id=5073&type=spkr&cid=0&srchText=IBM

7. Thompson, L. (2012, April) Using Data to Drive Sustainability. Paper contributed to the OSIsoft

User Conference Day 0 Sustainability Workshop, San Francisco, CA. To be posted on

www.osisoft.com.

8. Jones, M. (2008) PI Infrastructure for High Value Projects. Paper contributed to the OSIsoft

Regional Seminar, Kansas City, KA. Retrieved from http://www.osisoft.com/Templates/item-

abstract.aspx?id=1517&type=spkr&cid=0&srchText=Cargill

9. Jones, M. (2008) PI Infrastructure for High Value Projects. Paper contributed to the OSIsoft

Regional Seminar, Columbus, OH. Retrieved from http://www.osisoft.com/Templates/item-

abstract.aspx?id=1473&type=spkr&cid=0&srchText=Cargill

10. Heere, M. (2003) Getting ALL your data in PI. Paper contributed to the OSIsoft Transmission and

Distribution Conference West, San Francisco, CA. Retrieved from

http://www.osisoft.com/Templates/item-

abstract.aspx?id=1752&type=spkr&cid=0&srchText=PJM.

http://www.nytimes.com/2012/02/12/sunday-review/big-datas-impact-in-the-world.html?pagewanted=all

http://www.nytimes.com/2012/02/12/sunday-review/big-datas-impact-in-the-world.html?pagewanted=all

http://www.osisoft.com/Templates/item-abstract.aspx?id=5059

http://www.osisoft.com/Templates/item-abstract.aspx?id=5059

http://www.osisoft.com/Templates/item-abstract.aspx?id=5073&type=spkr&cid=0&srchText=IBM

http://www.osisoft.com/Templates/item-abstract.aspx?id=5073&type=spkr&cid=0&srchText=IBM


http://www.osisoft.com/Templates/item-abstract.aspx?id=1517&type=spkr&cid=0&srchText=Cargill




http://www.osisoft.com/Templates/item-abstract.aspx?id=1752&type=spkr&cid=0&srchText=PJM

http://www.osisoft.com/Templates/item-abstract.aspx?id=1752&type=spkr&cid=0&srchText=PJM



11. Lee,P. (2009, April) Decision & Operational Support in an Integrated Energy Business. Paper

contributed to the OSIsoft User Conference, San Francisco, CA. Retrieved from

http://www.osisoft.com/Templates/item-

abstract.aspx?id=4000&type=spkr&cid=0&srchText=Sempra

12. Hanson, J (2010, April). OSIsoft’s Enterprise PI as an Enabler for an Integrated Manufacturing

Landscape. Paper contributed to the OSIsoft User Conference, San Francisco, CA. Retrieved

from http://www.osisoft.com/Templates/item-

abstract.aspx?id=5065&type=indust&cid=0&industry=Pharmaceuticals Food $ Life

Science&value=Value (all)&tab=0&dindex=-1

13. Howard, A. (2007, April). The Considerations and Architecture for One of the Industry’s Most

Integrated and Automated MES Programs. Paper contributed to the OSIsoft User Conference,

Monterey, CA. Retrieved from http://www.osisoft.com/Templates/item-

abstract.aspx?id=2021&type=spkr&cid=0&srchText=Bristol Myers Squibb.

14. Reddy, M. (2006, April). Journey to a Real-time Enterprise. Paper contributed to the OSIsoft User

Conference, San Francisco, CA. Retrieved from http://www.osisoft.com/Templates/item-

abstract.aspx?id=2091&type=spkr&cid=0&srchText=chevron

http://www.osisoft.com/Templates/item-abstract.aspx?id=4000&type=spkr&cid=0&srchText=Sempra

http://www.osisoft.com/Templates/item-abstract.aspx?id=4000&type=spkr&cid=0&srchText=Sempra

http://www.osisoft.com/Templates/item-abstract.aspx?id=2091&type=spkr&cid=0&srchText=chevron

http://www.osisoft.com/Templates/item-abstract.aspx?id=2091&type=spkr&cid=0&srchText=chevron



ABOUT OSISOFT, LLC

OSIsoft (www.osisoft.com) delivers the PI System, the industry standard in enterprise infrastructure, for management of real-time data and events. With installations of the PI System in 110 countries spanning the globe, the OSIsoft PI System is used in manufacturing, energy, utilities, life sciences, data centers, facilities, and the process industries. This global installed base relies upon the OSIsoft PI System to safeguard data and deliver enterprise-wide visibility into operational, manufacturing and business data. The PI System enables users to manage assets, mitigate risks, comply with regulations, improve processes, drive innovation, make business decisions in real time, and to identify competitive business and market opportunities.

Founded in 1980, OSIsoft, LLC is headquartered in San Leandro, CA (with operations worldwide) and is privately held. Learn more about OSIsoft and the PI System at www.osisoft.com. ©1992-2011 OSIsoft, Inc. All rights reserved.


Documents

Streaming Data and Event Infrastructure - OSIsoftcdn.osisoft.com/corp/en/docs/whitepapers/WP... · Streaming Data and Event Infrastructure Last Saved May 4, 2012 2 of 20 ... However,