SIGHT MACHINE WHITE PAPER Manufacturing Analytics: Uncovering Secrets on Your Factory Floor Quick Take For manufacturers, operational insight is often masked by mountains of process and part data flowing from factory floor sensors, automation systems, suppliers, and other data sources. In effect, the factory floor holds secrets related to data correlations, trends, anomalies/deviations, root causes, and other key operational information. This lack of insight leads to lower quality rates, production inefficiencies, and lack of trust in the limited data that is available. True insight requires real-time reporting and deep dive analytics of historical trends; these capabilities reveal production and supply chain patterns, anomalies, quality and root cause issues, and other meaningful trends in time for the operations and quality teams to take effective actions. This paper makes the case for an advanced manufacturing analytics platform that helps uncover secrets from the factory floor for real-time, data-driven insight, enabling fact-based decisions.
2 Factory Floor Secrets Today s manufacturer lacks real-time analysis from the factory floor. While every machine, scanner, and sensor puts out a steady stream of structured or unstructured data, companies can t easily make sense of the volume, variety and velocity of data locked behind hard-to-use tools. Typical refrains from manufacturers include: - Our systems collect machine downtime metrics and other production data, but we don t have the means to fully understand what s important, or which metrics tell the full story of our operations in real time. In a way we re working blind. We don t have the ability to take corrective actions. - We have a scrap rate of 30%. We believe the problem is somewhere in the casting operation but we can t detect deviations and correlate results. What s missing is our ability to use the scrap data to find the root cause, and take action based on real-time information. - Our operations manager walks from machine to machine with a thumb drive to gather the data he needs, then spends hours working on a spreadsheet to correlate and analyze the data manually, before finally distributing the output to teammates. This isn t workable. By the time we get the data, it s too late to fix what was broken. In each example, the problem isn t lack of data. The challenge is meaningful and real-time insight into that data. As one manufacturer put it, It s almost as if our factory data holds secrets about key performance trends that we just can t get at. Data Rich, Information Poor Many, if not most, manufacturers have a set of tracking tools in place that monitor some aspects of production. Looking closer, legacy systems typically track: Process Data such as the rate at which parts are produced, when machines are online/offline, delivery/ on-time shipping, actual vs. planned production hours, OEE, etc. Parts Data number of parts produced and quality figures including the percentage meeting requirements, scrap, rework, warranty, etc. While process and parts data are sometimes captured, insight is not readily available in real time, nor interpreted with context or meaning. For example, drilling down from a downtime report to find out WHY a
3 system is down: Is it a lack of materials? Machine malfunction? Missing personnel? How fast is the problem corrected? What are the trends for the various problems? And what are the root causes? The industry has coined the acronym DRIP Data Rich, Information Poor to describe this common scenario. DRIP results in significant business challenges such as lower quality rates, lack of visibility into product efficiency, and lack of trust in the data that s available. Timing is also a challenge. Downtime and other metrics might be tracked with a variety of systems, but by the time the company accesses machine downtime metrics, it s too late to address the underlying problem. Again, the issue is not in the seeing of factory data. It s the lack of insight, the ability to get actionable information at the right time. Advanced analytics refers to the application of statistics and other mathematical tools to business data in order to assess and improve practices. In manufacturing, operations managers can use advanced analytics to take a deep dive into historical process data, identify patterns and relationships among discrete process steps and inputs, and then optimize the factors that prove to have the greatest effect on yield. Many global manufacturers in a range of industries and geographies now have an abundance of real-time shop floor data and the capability to conduct such sophisticated statistical assessments. They are taking previously isolated data sets, aggregating them, and analyzing them to reveal important insights. An integrated approach to data collection, presentation, and analysis has significant business value. With more intelligent use of the steady stream of factory data, today s manufacturer is in a better competitive position. Timing is also a challenge. Downtime and other metrics might be tracked with a variety of systems, but by the time the company accesses machine downtime metrics, it s too late to address the underlying problem. Again, the issue is not in the seeing of factory data. It s the lack of insight, the ability to get actionable information at the right time. Uncovering Secrets of the Factory Floor To overcome the Data Rich, Information Poor challenge, manufacturers need an advanced platform to make sense of those secrets masked by big data without the need to rip and replace installed systems. A recent article published by the Academy of Mechanical Engineers 1 emphasizes the benefits of harnessing the constant flow of factory floor big data: Big Data requires management tools to make sense of large sets of heterogeneous information. In the case of a factory, sources of data include CAD models, sensors, instruments, Internet transactions, simulations potentially, records of all the digital sources of information in the enterprise. The data bank is large, complex, and often fast moving, and so it becomes difficult to process using traditional database analysis and management tools. Industry stands to reap many benefits from Big Data as more sophisticated and automated data analytics technologies are developed. These technologies will help extract value and hidden knowledge from large, diverse data streams. It s important to note that the approach does not require removing current information tools such as
4 sensors, actuators, computerized controls, production management software, ERP and the like. These systems are already in place tracking specific forms of data that all need to be fed into the analytics platform. The best approach is to enhance these systems with a manufacturing analytics platform that lies on top of current systems already in place, without requiring a rip and replace implementation. The new systems must be able to extract value and knowledge, display trends and alert to anomalies and deviations, all without any disruptions to the manufacturing operation. In other words, a manufacturing analytics platform should maximize the ROI of the systems that you already have in place. Advanced Manufacturing Analytics Uncovering the secrets of the factory floor requires manufacturing analytics technology to effectively aggregate and analyze the incredible volume, velocity, and variety of data that is being created on your factory floor. Advanced manufacturing analytics help extract value and hidden knowledge from large, diverse data streams. The analytics platform must be a cloud solution built on a modern web framework with reports, charts, and dashboards running within a standard web browser on any device, from a PC or tablet to a mobile device, with access to real-time information from the factory floor anywhere, anytime. As one example of this approach, manufacturing analytics help provide insight into OEE. Typically, this metric is composed of three key areas, all of which interact with each other, to provide an overview of how a machine, a production line, or a factory is operating. Quality measuring how many parts were produced, and the percentage of those parts that meet quality requirements. Performance displaying the rate at which parts were produced when the machine, line or factory was in production. Availability showing actual vs. planned production hours. A manufacturing analytics platform reveals not only these three key areas, but also various statistics, such as the longest periods of downtime, systems that are particularly problematic in terms of root cause, or work shifts that suffer the lowest productivity. A graphical report groups and charts downtime by reason, so a company can allocate resources to tackle challenges in order of priority in real time. An advanced manufacturing analytics platform offers the same views and analysis for quality and performance. The user easily sees various events sorted by reason, which is extremely useful for controls engineering to investigate root cause. By showing downtime details, an engineer sees what happened to the machine
5 right before it suffered an outage, then can jump to the corresponding cycle. The engineer scans the data, diagnoses the problem, and dispatches maintenance to fix the issue. All of this happens in real time, without waiting for teams to research the problem and design experiments to try and reproduce the problem. The approach drastically reduces the time associated with diagnosing a problem, improves plant and machine efficiency levels, and improves quality by guiding managers toward problem resolution, from anywhere in the world. In addition, manufacturing analytics platforms provide the tools needed to run deep dive analysis of hundreds even thousands of variables in databases that are multiple terabytes in size. Sorting through the data to identify bottlenecks in a line or factory, for example, can take months with legacy tools, but can be accomplished in a fraction of the time using a manufacturing analytics solution. Final Thoughts In all, by analyzing and presenting data for actionable use, companies improve business decisions. The Sight Machine manufacturing analytics platform was designed with this goal in mind. The manufacturing analytics platform reveals patterns, trends, and data correlations and flags issues. The solution was designed to cover the entire spectrum of data analysis needs, so manufacturers can move effortlessly from data to understanding. With it, manufacturers can uncover the secrets hidden on the factory floor.
6 Reference: 1 Putting Big Data to Work Ahmed K. Noor 2014 American Society of Mechanical Engineers (ASME) https://www.asme.org/wwwasmeorg/media/resourcefiles/network/media/mechanical%20engineering%20 Magazine/1013BigData.pdf About Sight Machine The Sight Machine Manufacturing Analytics Platform analyzes existing manufacturing data for trends and important statistics, presenting real-time information to manufacturers in an easily digestible cloud-based format. Integrated into existing manufacturing operations, Sight Machine applies best manufacturing practices, such as standardized adapters for legacy MES, ERP and SCADA systems, along with advanced data management such as signal processing of sensors and images to higher-level data structures. Sight Machine s web and mobile-based applications are available on demand, at any time, from virtually any connected device from across the plant or across the globe.