Skip to content

Posts tagged ‘Digital Manufacturing’

Five Ways AI Can Help Create New Smart Manufacturing Startups

smart manufacturing, AI, machine learning

AI and machine learning’s potential to drive greater visibility, control, and insight across shop floors while monitoring machines and processes in real-time continue to attract venture capital. $62 billion is now invested in 5,396 startups concentrating on the intersection of AI, machine learning, manufacturing, and Industry 4.0, according to Crunchbase.

PwC’s broader tech sector analysis shows a 30% year-over-year growth in funding rounds that reached $293.2 billion in 2021. Smart manufacturing startups are financed by seed rounds at 52%, followed by early-stage venture funding at 33%. The median last funding amount was $1.6 million, with the average being $9.93 million.

 Abundant AI startup opportunities in smart manufacturing and industry 4.0 

According to Gartner, “The underlying concept of Industry 4.0 is to connect embedded systems and smart production facilities to generate a digital convergence between industry, business, and internal functions and processes.” As a result, Industry 4.0 is predicted to grow from $84.59 billion in 2020 to $334.18 billion by 2028. AI and machine learning adoption in manufacturing are growing in five core fields: smart production, products and services improvements, business operations and management, supply chain, and business model decision-making. Deloitte’s survey on AI adoption in manufacturing found that 93% of companies believe AI will be a key technology to drive growth and innovation.

Machine intelligence (MI) is one of the primary catalysts driving increased venture capital investment in smart manufacturing. Startup CEOs and their customers want AI and machine learning models based on actual data, and machine intelligence is helping to make that happen. An article by McKinsey & Company provides valuable insights into market gaps for new ventures. McKinsey’s compelling data point is that those leading companies using MI achieve 3X to 4X the impact of their peers. However, 92% of leaders also have a process to track incomplete or inaccurate data – which is another market gap startups need to fill.

AI, Industry 4, smart manufacturing

McKinsey and the Massachusetts Institute of Technology (MIT) collaborated on a survey to identify machine intelligence leaders’ KPI gains relative to their peers. They found that leaders achieve efficiency, cost, revenue, service, and time-to-market advantages. Source: Toward smart production: Machine intelligence in business operations, McKinsey & Company. February 1, 2022.

Based on the uplift MI creates for new smart manufacturing startup funding and the pervasive need manufacturers have to improve visibility & control across shop floors, startups have many potential opportunities. The following are five that AI and machine learning is helping to create:

  1. AI-enabled Configure, Price, and Quote (CPQ) systems that can factor in supply chain volatility on product costs are needed. Several startups are already using AI and machine learning in CPQ workflows, and they compete with the largest enterprise software providers in the industry, including Salesforce, SAP, Microsoft, and others. However, no one has taken on the challenge of using AI to factor in how supply chain volatility changes standard and actual costs in real-time. For example, knowing the impact of pricing changes based on an allocation, how does that impact standard costs per unit on each order? Right now, an analyst needs to spend time doing that. AI and machine learning could take on that task so analysts could get to the larger, more complex, and costly supply chain problems impacting CPQ close rates and revenue.
  2. Using AI-enabled real-time data capture techniques to identify anomalies in throughput as an indicator of machine health. The aggregated data manufacturing operations produced every day holds clues regarding each machine’s health on the shop floor. Automated data capture can identify scrap rates, yield rates and track actual costs. However, none of them can analyze the slight variations in process flow product outputs to warn of possible machine or supply chain issues. Each process manufacturing machine runs at its cadence or speed, and having an AI-based sensor system track and analyze why speeds are off could save thousands of dollars in maintenance costs and keep the line running. In addition, adding insight and intelligence to the machine’s real-time data feeds frees quality engineers to concentrate on more complex problems.
  3. Industrial Internet of Things (IIoT) and edge computing data can be used for fine-tuning finite scheduling in real-time. Finite scheduling is part of the broader manufacturing systems organizations rely on to optimize shop floor schedules, machinery, and staff scheduling. It can be either manually intensive or automated to provide operators with valuable insights. A potential smart manufacturing opportunity is a finite scheduler that relies on AI and machine learning to keep schedules on track and make trade-offs to ensure resources are used efficiently. Finite schedulers also need greater accuracy in factoring in frequent changes to delivery dates. AI and machine learning could drive greater on-time delivery performance when integrated across all the shop floors a manufacturer relies on.
  4. Automated visual inspections and quality analysis to improve yield rates and reduce scrap. Using visual sensors to capture data in real-time and then analyze them for anomalies is in its nascent stages of deployment and growth. However, this is an area where captured data sets can provide machine learning algorithms with enough accuracy to identify potential quality problems on products before they leave the factory. Convolutional neural networks are an effective machine learning technique for identifying patterns and anomalies in images. They’re perfect for the use case of streamlining visual inspection and in-line quality checks in discrete, batch, and process manufacturing.
  5. Coordinated robotics (Cobots) to handle assemble-to-order product assembly. The latest cobots can be programmed to stay in sync with each other and perform pick, pack, ship, and place materials in warehouses. What’s needed are advanced cobots that can handle simple product assembly at a more competitive cost as manufacturers continue to face chronic labor shortages and often run a shift with less than half the teams they need.

Talent remains an area of need 

Manufacturers’ CEOs and COOs say that recruiting and retaining enough talent to run all the production shifts they need is the most persistent issue. In addition, those manufacturers located in remote regions of the world are turning to robotics to fulfill orders, which opens up opportunities for integrating AI and machine learning to enable cobots to complete assemble-to-order tasks. The unknown impact of how fast supply chain conditions change needs work from startups, too, especially in tracking actual cost performance. These are just a few opportunities for startups looking to apply AI and machine learnings’ innate strengths to solve complex supply chain, manufacturing, quality management, and compliance challenges.

10 Ways Machine Learning Is Revolutionizing Manufacturing In 2018

  • Improving semiconductor manufacturing yields up to 30%, reducing scrap rates, and optimizing fab operations is achievable with machine learning.
  • Reducing supply chain forecasting errors by 50% and lost sales by 65% with better product availability is achievable with machine learning.
  • Automating quality testing using machine learning is increasing defect detection rates up to 90%.

Bottom line: Machine learning algorithms, applications, and platforms are helping manufacturers find new business models, fine-tune product quality, and optimize manufacturing operations to the shop floor level.

Manufacturers care most about finding new ways to grow, excel at product quality while still being able to take on short lead-time production runs from customers. New business models often bring the paradox of new product lines that strain existing ERP, CRM and PLM systems by the need always to improve time-to-customer performance. New products are proliferating in manufacturing today, and delivery windows are tightening. Manufacturers are turning to machine learning to improve the end-to-end performance of their operations and find a performance-based solution to this paradox.

The ten ways machine learning is revolutionizing manufacturing in 2018 include the following:

  • Improving semiconductor manufacturing yields up to 30%, reducing scrap rates, and optimizing fab operations are is achievable with machine learning. Attaining up to a 30% reduction in yield detraction in semiconductor manufacturing, reducing scrap rates based on machine learning-based root-cause analysis and reducing testing costs using AI optimization are the top three areas where machine learning will improve semiconductor manufacturing. McKinsey also found that AI-enhanced predictive maintenance of industrial equipment will generate a 10% reduction in annual maintenance costs, up to a 20% downtime reduction and 25% reduction in inspection costs. Source: Smartening up with Artificial Intelligence (AI) – What’s in it for Germany and its Industrial Sector? (52 pp., PDF, no opt-in) McKinsey & Company.

  • Asset Management, Supply Chain Management, and Inventory Management are the hottest areas of artificial intelligence, machine learning and IoT adoption in manufacturing today. The World Economic Forum (WEF) and A.T. Kearney’s recent study of the future of production find that manufacturers are evaluating how combining emerging technologies including IoT, AI, and machine learning can improve asset tracking accuracy, supply chain visibility, and inventory optimization. Source: Technology and Innovation for the Future of Production: Accelerating Value Creation (38 pp., PDF, no opt-in) World Economic Forum with A.T. Kearney.

  • Manufacturer’s adoption of machine learning and analytics to improve predictive maintenance is predicted to increase 38% in the next five years according to PwC. Analytics and MI-driven process and quality optimization are predicted to grow 35% and process visualization and automation, 34%. PwC sees the integration of analytics, APIs and big data contributing to a 31% growth rate for connected factories in the next five years. Source: Digital Factories 2020: Shaping the future of manufacturing (48 pp., PDF, no opt-in) PriceWaterhouseCoopers

  • McKinsey predicts machine learning will reduce supply chain forecasting errors by 50% and reduce lost sales by 65% with better product availability. Supply chains are the lifeblood of any manufacturing business. Machine learning is predicted to reduce costs related to transport and warehousing and supply chain administration by 5 to 10% and 25 to 40%, respectively. Due to machine learning, overall inventory reductions of 20 to 50% are possible. Source: Smartening up with Artificial Intelligence (AI) – What’s in it for Germany and its Industrial Sector? (52 pp., PDF, no opt-in) McKinsey & Company.

  • Improving demand forecast accuracy to reduce energy costs and negative price variances using machine learning uncovers price elasticity and price sensitivity as well. Honeywell is integrating AI and machine-learning algorithms into procurement, strategic sourcing and cost management. Source: Honeywell Connected Plant: Analytics and Beyond. (23 pp., PDF, no opt-in) 2017 Honeywell User’s Group.

  • Automating inventory optimization using machine learning has improved service levels by 16% while simultaneously increasing inventory turns by 25%. AI and machine learning constraint-based algorithms and modeling are making it possible scale inventory optimization across all distribution locations, taking into account external, independent variables that affect demand and time-to-customer delivery performance. Source: Transform the manufacturing supply chain with Multi-Echelon inventory optimization, Microsoft, March 1, 2018.

  • Combining real-time monitoring and machine learning is optimizing shop floor operations, providing insights into machine-level loads and production schedule performance. Knowing in real-time how each machine’s load level impacts overall production schedule performance leads to better decisions managing each production run. Optimizing the best possible set of machines for a given production run is now possible using machine learning algorithms. Source: Factories of the Future: How Symbiotic Production Systems, Real-Time Production Monitoring, Edge Analytics and AI Are Making Factories Intelligent and Agile, (43 pp., PDF, no opt-in) Youichi Nonaka, Senior Chief Researcher, Hitachi R&D Group and Sudhanshu Gaur Director, Global Center for Social Innovation Hitachi America R&D

  • Improving the accuracy of detecting costs of performance degradation across multiple manufacturing scenarios reduces costs by 50% or more. Using real-time monitoring technologies to create accurate data sets that capture pricing, inventory velocity, and related variables gives machine learning apps what they need to determine cost behaviors across multiple manufacturing scenarios. Source: Leveraging AI for Industrial IoT (27 pp., PDF, no opt-in) Chetan Gupta, Ph.D. Chief Data Scientist, Big Data Lab, Hitachi America Ltd. Date: Sept. 19th, 2017

  • A manufacturer was able to achieve a 35% reduction in test and calibration time via accurate prediction of calibration and test results using machine learning. The project’s goal was to reduce test and calibration time in the production of mobile hydraulic pumps. The methodology focused on using a series of machine learning models that would predict test outcomes and learn over time. The process workflow below was able to isolate the bottlenecks, streamlining test and calibration time in the process. Source: The Value Of Data Science Standards In Manufacturing Analytics (13 pp., PDF, no opt-in) Soundar Srinivasan, Bosch Data Mining Solutions And Services

  • Improving yield rates, preventative maintenance accuracy and workloads by the asset is now possible by combining machine learning and Overall Equipment Effectiveness (OEE). OEE is a pervasively used metric in manufacturing as it combines availability, performance, and quality, defining production effectiveness. Combined with other metrics, it’s possible to find the factors that impact manufacturing performance the most and least. Integrating OEE and other datasets in machine learning models that learn quickly through iteration are one of the fastest growing areas of manufacturing intelligence and analytics today. Source: TIBCO Manufacturing Solutions, TIBCO Community, January 30, 2018

Additional reading:

Artificial Intelligence (AI) Delivering Breakthroughs in Industrial IoT (26 pp., PDF, no opt-in) Hitachi

Artificial Intelligence and Robotics and Their Impact on the Workplace (120 pp., PDF, no opt-in) IBA Global Employment Institute

Artificial Intelligence: The Next Digital Frontier? (80 pp., PDF, no opt-in) McKinsey and Company

Big Data Analytics for Smart Manufacturing: Case Studies in Semiconductor Manufacturing (20 pp., PDF, no opt-in), Applied Materials, Applied Global Services

Connected Factory and Digital Manufacturing: A Competitive Advantage, Shantanu Rai, HCL Technologies (36 pp., PDF, no opt-in)

Demystifying AI, Machine Learning, and Deep Learning, DZone, AI Zone

Digital Factories 2020: Shaping the future of manufacturing (48 pp., PDF, no opt-in) PriceWaterhouseCoopers

Emerging trends in global advanced manufacturing: Challenges, Opportunities, And Policy Responses (76 pp., PDF, no opt-in) University of Cambridge

Factories of the Future: How Symbiotic Production Systems, Real-Time Production Monitoring, Edge Analytics and AI Are Making Factories Intelligent and Agile, (43 pp., PDF, no opt-in) Youichi Nonaka, Senior Chief Researcher, Hitachi R&D Group and Sudhanshu Gaur Director, Global Center for Social Innovation Hitachi America R&D

Get started with the Connected factory preconfigured solution, Microsoft Azure

Honeywell Connected Plant: Analytics and Beyond. (23 pp., PDF, no opt-in) 2017 Honeywell User’s Group.

Impact of the Fourth Industrial Revolution on Supply Chains (22 pp., PDF, no opt-in) World Economic Forum

Leveraging AI for Industrial IoT (27 pp., PDF, no opt-in) Chetan Gupta, Ph.D. Chief Data Scientist, Big Data Lab, Hitachi America Ltd. Date: Sept. 19th, 2017

Machine Learning & Artificial Intelligence Presentation (14 pp., PDF, no opt-in) Erik Hjerpe Volvo Car Group

Machine Learning Techniques in Manufacturing Applications & Caveats, (44 pp., PDF, no opt-in), Thomas Hill, Ph.D. | Exec. Director Analytics, Dell

Machine learning: the power and promise of computers that learn by example (128 pp., PDF, no opt-in) Royal Society UK

Predictive maintenance and the smart factory (8 pp., PDF, no opt-in) Deloitte

Priore, P., Gómez, A., Pino, R., & Rosillo, R. (2014). Dynamic scheduling of manufacturing systems using machine learning: An updated reviewAi Edam28(1), 83-97.

Smartening up with Artificial Intelligence (AI) – What’s in it for Germany and its Industrial Sector? (52 pp., PDF, no opt-in) McKinsey & Company

Technology and Innovation for the Future of Production: Accelerating Value Creation (38 pp., PDF, no opt-in) World Economic Forum with A.T. Kearney

The Future of Manufacturing; Making things in a changing world (52 pp., PDF, no opt-in) Deloitte University Press

The transformative potential of AI in the manufacturing industry, Microsoft, by Sanjay Ravi, Managing Director, Worldwide Discrete Manufacturing, Microsoft, September 25, 2017

The Value Of Data Science Standards In Manufacturing Analytics (13 pp., PDF, no opt-in) Soundar Srinivasan, Bosch Data Mining Solutions And Services

TIBCO Manufacturing Solutions, TIBCO Community, January 30, 2018

Transform the manufacturing supply chain with Multi-Echelon inventory optimization, Microsoft, March 1, 2018.

Turning AI into concrete value: the successful implementers’ toolkit (28 pp., PDF, no opt-in) Capgemini Consulting

Wuest, T., Weimer, D., Irgens, C., & Thoben, K. D. (2016). Machine learning in manufacturing: advantages, challenges, and applicationsProduction & Manufacturing Research4(1), 23-45.

10 Ways Analytics Are Accelerating Digital Manufacturing

  • 42% of manufacturers say big data and analytics as their highest priority in 2015.
  • 56% of power distribution providers rank big data and analytics within their top three priorities for 2015.
  • 61% of aviation companies consider big data and analytics their highest priority this year.

Bottom line: Digital manufacturing strategies are gaining ground as manufacturers adopt big data and analytics to improve operational effectiveness, time-to-market, new product development and increase product quality and reliability.

Analytics Are Fueling Digital Manufacturing Growth

Big data and analytics adoption by manufacturers is the first step many are taking to create a galvanized, intelligent digital thread that unifies every aspect of their value chains. For aerospace manufacturers whose supply chains are exceptionally complex, big data and analytics are revolutionizing value chains starting with suppliers and progressing through all operations.

The majority of manufacturers are relying on analytics to improve order accuracy, shipment & cycle time performance, and product quality. Those excelling at digital manufacturing strategies are gaining additional analytical insights into how they can make decisions more accurately, quicker and with lower potential costs and risks.

The manufacturing industry generates more data than any other sector of the global economy on a consistent basis.   The more complex a given manufacturers’ operations are, the more valuable the insights gained from big data and analytics. The following comparison of big data analytics priorities by industry from a recent speech given by Jeff Immelt, CEO and President of General Electric illustrates this point:

analytics customer survey

Source: GE Minds and Machines Presentation, Jeff Immelt, CEO & President, General Electric.

10 Ways Analytics Are Accelerating Digital Manufacturing 

The ten ways analytics is accelerating digital manufacturing adoption globally include the following:

  • Providing real-time operator intelligence (70%), remote monitoring and diagnostics (66%), and condition-based maintenance (59%) are the three most valuable areas for analytics GE customers mentioned in a recent survey. GE’s industrial customers are looking to tailor pre-built applications that can deliver the eight different functional areas shown in the graphic below.  Manufacturers are looking to asset performance management as an integral part of their digital thread’s analytics and insight.

industrial customer perspective

Source: GE Minds and Machines Presentation, Jeff Immelt, CEO & President, General Electric.

  • Using data modeling to improve production workflows is improving Earnings Before Interest & Taxes (EBIT) by 55% for a chemical manufacturer.  Using analytics and data modeling to make more accurate,  efficient decisions encompassing making or buying ingredients, choosing to substitute an ingredient or not, optimizing equipment usage and/or reliability and gaining incremental sales through increased production capacity is leading to a significant improvement in EBIT for a leading chemical manufacturer on a consistent basis.  The following graphic provides insights into the contributions of each factor in improving EBIT performance.

EBIT Growth

Source: Taming manufacturing complexity with advanced analytics. McKinsey & Company by Patrick Briest, Valerio Dilda, and Ken Somers February 2015. 

  • Planning-execution integration in production centers and real-time production integration are two areas where analytics are having the greatest impact on manufacturers’ operating expenses (OPEX). When analytics are integrated as part of a digital manufacturing strategy, supply chains benefit when Web-EDI (Electronic Data Interchange) and real-time order conformation are implemented and analyzed for continual improvement.

Digital initaitves impact

Source: Operational Excellence through Digital in Manufacturing Industries. Capgemini Consulting.

  • Optimization tools (56%), demand forecasting (53%), integrated business planning (48%) and supplier collaboration & risk analytics (46%) are being rapidly adopted by manufacturers today, setting the foundation for digital manufacturing growth.  Deloitte recently interviewed supply chain executives regarding the thirteen fastest-moving technical capacities they are using today and expect to use in the future. The following graphic provides an overview of supply chain capabilities current in use and what percent of each they expect to use in the future.

use of supply chain capabilities

Source: Supply Chain Talent of the Future Findings from the third annual supply chain survey. Deloitte.  2015.

  • Analytics is integral to making the vision of Industrie 4.0 a reality. Industrie 4.0 is a German government initiative that promotes automation of the manufacturing industry with the goal of developing Smart Factories. Analytics is extensively used in manufacturing centers who are in the process of reengineering their entire operations to attain Industrie 4.0 compliance. Manufacturing value chains in highly regulated industries that rely on German suppliers and manufacturers are also relying on analytics extensively to guide their Industrie 4.0 journey. A recent Deloitte study of Industrie 4.0 adoption found that research and development (43%) will see the greatest transformational contribution from Industry 4.0.

Industry 4.0 areas

Source: Industry 4.0: Challenges and solutions for the digital transformation and use of exponential technologies. Deloitte Consulting, 2015

  • Analytics is enabling manufacturers to also scale real time cloud-based operational intelligence, condition-based monitoring, monitoring & diagnostics and asset lifecycle management across global manufacturing centers.  Capturing, aggregating, analyzing and taking action on analytics across all production centers using the GE Predix Cloud will also accelerate digital manufacturing growth over time.  Integrating analytics, industrial and sensor data into a scalable series of data models and apps delivered as a Platform-as-a-Service (PaaS), GE will make this service commercially available in 2016.  The following graphic illustrates how complex manufacturers could use Predix Cloud to improve operational efficiency and quality.

horizontal capability controls

Source: Jeff Immelt Presentation on Pivot Strategy, December 16, 2014

  • Analytics is providing greater insights into product, process, program and service quality, forcing manufacturers to revamp existing production centers and make them more efficient.  Gaining greater insight into which production centers and factories are delivering the highest quality products and why is now possible.  The vision of unifying quality across an enterprise quality management and compliance (ECQM) framework is now a reality, driving greater digital manufacturing growth as a result. The following graphic from Tableau is an example of a manufacturing quality dashboard.

Mfg quality dashboard

Source: Manufacturing Analytics Quality Dashboard

  • Increasing production yields through the use of more effective supplier quality management and bill of material (BOM) planning integrated within production processes.  Analytics is extensively being used today for supplier audits, supplier quality management and traceability. Capitalizing on the full value of these analytics is a strong catalyst for manufacturers to move closer to digitizing their operations. Check out SelectHub’s Supply Chain Management (SCM) Buyer’s Guide and leaderboard to learn more about the SCM landscape.
  • Using analytics to predict machine failures before they occur reduces downtime, production costs and increase customer satisfaction.  In highly regulated industries production equipment is periodically audited and reviewed for conformance to specific standards.  Integrating even the simplest sensor into production equipment can deliver valuable insights into what factors cause it to fail.  Analytics are providing Failure Mode and Effects Analysis (FMEA) in real-time today, providing manufacturers with a glimpse into which equipment and machinery will most likely fail when. Knowing this can save literally millions of dollars in lost production time.
  • Adopting Pareto Analysis to continually improve schedule, quality and cost performance to the cell or production center level is driving digital manufacturing adoption.  Determining which factors are enhancing or reducing product, process and program quality is now possible using advanced manufacturing analytics. Differentiating between the many symptoms of a quality problem and its root cause is now becoming possible, especially for companies pursuing digital manufacturing strategies.

Additional sources of information on the impact of analytics on digital manufacturing:

 

%d bloggers like this: