Manufacturing Engineering: Machine monitoring and analytics have become more critical than ever in the digital manufacturing/Industry 4.0 era. What’s new with Sight Machine’s latest offerings?
Jon Sobel: Our latest offerings are designed for a far broader range of manufacturers. While the majority of manufacturers have at least begun their digital transformation journeys, we’ve found a wide spectrum of readiness levels among them. A lot of companies are eager to use IoT [Internet of Things] data to get better visibility into their manufacturing operations, but aren’t ready for advanced analytics. They might not have the in-house expertise, the organizational support, or it represents too big a leap to take in one step. We’ve introduced a new tier of our digital manufacturing platform: Enterprise Manufacturing Visibility [EMV], which offers a rapid entry into digital manufacturing. EMV lets companies very quickly gain real-time visibility into the performance of their manufacturing operations across all their plants. It offers self-service setup, enabling manufacturers to quickly pull in all kinds of production data from factory floors and then monitor that data through browser-based visibility.
Sight Machine’s Enterprise Manufacturing Analytics [EMA], our advanced tier, offers a full suite of analytic tools like statistical process control, correlation heat maps, anomaly detection and bottleneck analysis. It applies advanced analytics to achieve benefits like increased production, reduced cycle times, and decreased scrap rates. These analytics-based insights are the foundation needed for companies to transform their business models and their relationships with suppliers and customers.
ME: How can your new tiered structure help enable manufacturers to get started in digital manufacturing without doing full-fledged analytics?
Sobel: Until now, we had a single offering that delivered both visibility and analytics. Now, we have broken our visibility tools out into a stand-alone product. EMV provides a straightforward, self-service, and easily scalable entry into digital manufacturing, with automated tools for ingesting the data, cleaning and contextualizing it, and modeling the production process.
EMV lets companies monitor output, availability and downtime by facility, machine type and machine through applications that include Global Operations View, KPI Dashboard and Streaming Data Visualization. The visibility is both real-time and historical.
EMV’s features include Digital Twin Builder, our browser-based tool to create data models of facilities, machines and machine types. It also includes FactoryTX, our edge software that prepares machine and factory data for streaming to the cloud.
ME: How difficult is it for companies to get started with monitoring/analytics and what does your recently introduced Digital Readiness Index methodology tell them?
Sobel: We have found that the key to successful digital transformation is to pick projects appropriate for the readiness level of a company or plant. Readiness includes not just technical factors like data connectivity and accessibility, as well as cloud and security strategy, but also organizational factors. We’ve found in our six years working with G500 manufacturers that organizational factors are at least as important as technical factors. They include the level of commitment and buy-in both at the plant level and at the executive level, and the existence of cloud and security strategies.
Sight Machine has developed a methodology called the Digital Readiness Index [DRI] to evaluate a company’s technical and organizational readiness for digital manufacturing projects, and to identify appropriate projects—those most likely to succeed based on current readiness. We use a standardized questionnaire to map each company into one of five Digital Readiness Zones: Connection, Visibility, Efficiency, Advanced Analytics and Transformation. Each Digital Readiness Zone maps to examples of projects that are achievable for that level of readiness.
ME: At this point, how far along is the manufacturing industry in adopting advanced factory analytics?
Sobel: All the major manufacturers we speak with have at least begun the digital transformation journey. We find a wide variance not only among companies but within companies’ individual plants.
You can think about it in terms of these steps: data access; visibility; and insights. It is fairly common for companies to be collecting and storing their digital data. Many have cobbled together simple dashboards that allow them to monitor the different lines or machines within individual factories.
Relatively few manufacturers have gained visibility across their factories. We believe Sight Machine offers the only scalable solution on the market for rapidly gaining cross-enterprise visibility.
For advanced analytics, setting aside the companies we work with, we most often see bespoke data science projects. Companies facing a performance or quality issue will task their data scientists with solving the known problem. They will typically develop a one-off solution, taking whatever data they can get their hands on, throwing it into Excel or a statistical program, and looking for correlations.
The problem is that these projects are neither scalable nor applicable to other problems. The data extraction and modeling is done in a way that creates application and data silos. From the point of view of a VP of Operations or CIO, they’re left with a proliferation of customized applications that can’t be integrated in a manner that lends to understanding the manufacturing enterprise’s operations as a whole.
Sight Machine’s Digital Twin is a working, live data model of the production process. A single application that can be quickly expanded by adding new machines and processes, our Digital Twin provides a single source of truth that can be queried to get at any operational issue.
ME: What can advanced analytics offer manufacturers diving deep into this technology?
Sobel: The digital revolution has transformed sector after sector—advertising, retail, investing, science, politics—unleashing not only vast improvements in efficiency, but also transforming and creating new business models. Manufacturers are just beginning to tap into the power of their data and are seeing real results.
When you squeeze more productivity out of your existing plants and machines, or reduce scrap and defects, it directly impacts your bottom line. But most leading manufacturers picked all the low-hanging fruit long ago from disciplines like Six Sigma and Lean. The next big leaps will come from gaining true visibility into a manufacturer’s enterprise operations, across all factories, and then applying the analytical techniques honed over more than a decade from digital transformations in other industries.
ME: How does artificial intelligence [AI] in Sight Machine’s platform push factory analytics further than in previous or competing offerings?
Sobel: Sight Machine’s AI Data Pipeline is a patent-pending technology that transforms raw data into contextualized data to which analytics can be successfully applied. With our platform, the AI Data Pipeline does the grunt work of taking raw data from sensors, PLCs, data historians, etc., and then readying it for analysis via cleansing, tagging and blending.
The data then goes to our other patent-pending technology, the Plant Digital Twin, which models factory production processes. It takes the data from the AI Data Pipeline and assembles it into a model of the system, translating thousands of data points from hundreds of sources into a representation of parts or batches moving through production. The Plant Digital Twin reflects the machine state at each point in the production process, for each part or batch, and also records the raw material used for that part or batch, environmental factors like humidity and temperature, the identities of machine operators, and whatever additional data is available.
With competing offerings like in-house analytic initiatives, data scientists spend the majority of their time manually selecting, cleansing and combining data—not analyzing data to find actionable, business insights. By contrast, our AI Data Pipeline and Plant Digital Twin technologies automatically create a digital representation of production processes in real-time, leaving data scientists, line operators, and plant managers free to proactively respond to operations issues, rather than retroactively investigate them.
ME: Give me an example of a manufacturer doing exemplary work with your manufacturing analytics platform.
Sobel: We were brought in by a healthcare products manufacturer to help solve a problem that other technology providers were unable to crack—a high scrap rate in their most profitable plant. The Sight Machine platform performed root-cause analysis that blended natural language processing, sequence analysis, cluster analysis and regressions.
The Sight Machine platform identified and prioritized the multiple causes of scrap, leading to an increase in Overall Equipment Effectiveness of about 3%. This translates into potential savings of more than $20 million per year for this plant alone.
Siemens PLM Software (Plano, TX) has agreed to acquire Solido Design Automation Inc. (Saskatoon, SK), a developer of variation-aware design and characterization software to semiconductor companies.
Terms of the transaction were not disclosed. Siemens said it expected to close the transaction in early December 2017. Solido’s machine learning-based products are currently used in production at over 40 major companies, enabling them to design, verify, and manufacture more competitive products.
The acquisition further expands the analog/mixed-signal (AMS) verification portfolio of Mentor Graphics (acquired by Siemens in March 2017), enabling it to address the growing challenges of IC design and verification for automotive, communications, data-center computing, networking, mobile, and IoT applications.
Solido helps its customers to address the impact of variability to improve IC performance, power, area, and yield, said Amit Gupta, Solido Design Automation founder, president and CEO. “Combining our technology portfolio with Mentor’s IC capabilities and market reach will allow us to provide solutions to the semiconductor industry on an even larger scale,” Gupta said. “We are excited to contribute to Siemens’ broader digitalization strategy with our applied machine learning for engineering technology portfolio and expertise.”
Tebis America (Troy, MI) announced Release 5 of its Version 4.0 CAD/CAM software featuring optimized performance that helps users accelerate their processes.
Updates in this version allow users to speed processes without functional restrictions with improvements to machine simulation, working with tool sets, searching for tools in feature machining, or exchanging tools in the Job Manager.
With this new release, NC programming is now largely automated based on templates with process libraries that enable fast and reliable procedures and processes. Users can also edit large and complex parts with the software. Tebis identified the heaviest loads that occur in specific processes, helping alleviate bottlenecks that can result in long waiting times as well as heavy use of resources and conflicts. Tebis developers adapted the system to optimize the use of available memory, and multicore technology relying on parallel processing was integrated at the same time. The extended parallel processing saves significant time, especially in the calculation of NC programs for re-roughing. Parts can be loaded, shaded and saved with time optimization.
Manufacturing execution systems (MES) developer 42Q (San Jose, CA) has released its new Digital Factory Starter Kit, a solution with key capabilities that accelerate digital factory transformation.
The Digital Factory Starter Kit includes shop floor and quality functions, traceability capability and business intelligence (BI) reporting. Part of 42Q’s cloud solution, this kit enables manufacturers to realize value from digital factory transformation using 42Q in a few weeks.
“The costs and complexity of deploying a conventional on-premise MES platform can result in a time-to-value of between 9–18 months for basic functionality,” said Srivats Ramaswami, CTO of 42Q. “Manufacturers beginning their digital transformation should not have to wait that long. We designed the Digital Factory Starter Kit to include the core features necessary for digital transformation, and get companies up and running in a few weeks with no interruption to business processes or customer deliveries.”
Key capabilities of the Digital Factory Starter Kit include electronic travelers, cycle time management, shop order management and process
routing control. Functionality for quality management is fully integrated, including employee verification and electronic work instructions, traceability, product genealogy, labeling and control plans. The Digital Factory Starter Kit also includes 42Q’s Business Intelligence module, which provides tools for real time alerts from manufacturing operations along with data visualization.
The solution enables companies to manage quality and operations, monitor yields, WIP, and throughput in real time. “This solution is proven and already deployed in many manufacturing facilities in highly regulated industries, including medical, aerospace and automotive manufacturing,” Ramaswami said.
Computerized maintenance management software (CMMS) developer EZmaintain (Cleveland) has introduced its cloud-based CMMS package for integrating IoT sensors to measure temperature and vibration levels of equipment such as motors, gear boxes and generators.
Users can run the web-based CMMS independently to manage failure or preventive maintenance activities. Adding these smart IoT sensors allows users to capture real-time data for condition monitoring with alerts viewed directly via a CMMS dashboard.
EZmaintain CMMS works across different devices, from desktops to mobile devices. Users can add or configure Temperature/Vibration sensors via the dashboard, and can mount IoT sensors on various assets that need monitoring using screw-type mounting designs. With these low-cost, easy-to-setup sensors, the software can be used in various industrial or commercial applications.
Software Update is edited by Senior Editor Patrick Waurzyniak; firstname.lastname@example.org.