MANUFACTURING
Predictive analytics plays an important part in manufacturing, planning, warranty analysis, supply chain intelligence, and so on. Customers are demanding quality products at lower prices with faster delivery. Hence it has become vital for manufacturers to deliver a differentiated experience throughout the entire customer lifecycle. Through the implementation of predictive analytics, manufacturers can predict the tendency of a prospect to purchase and the level to which a prospect would be accepting the given offers. Additionally, by analyzing group behavior manufacturers can plan logistics and operations that are based on specific demand for a particular type of product in the given region.
With the advent of technologies such as Industrial Internet of Things (IIoT), the predictive analytics is becoming more intellectual empowering the manufacturing industry. Along with the rising demand, the manufactures are also using the capabilities of predictive analytics to enhance their production abilities. Countries such as US and China have put the use of predictive analytics for manufacturing as to improve the manufacturing at global scale. There will be prominent use of predictive analytics solutions in manufacturing vertical for predictive maintenance, asset management, and remote monitoring applications.
COMPETITIVE LEADERSHIP MAPPING TERMINOLOGY
The vendors of predictive analytics software in manufacturing are placed into 4 categories based on their performance in each criterion: “visionary leaders,” “innovators,” “dynamic differentiators,” and “emerging companies.” The top 24 vendors evaluated in the data quality tools market include Agilone, Alteryx, Inc, Angoss Software Corporation, Dataiku, Domino Data Lab, Exago, Inc., Fair Isaac Corporation (Fico), Good Data, Greenwave Systems, Inc, IBM Corporation, Information Builders, Inc., Knime Ag, Kognitio, Microsoft Corporation, NTT Data Corporation, Oracle Corporation, Qliktech, Inc., Rapidminer, Inc, SAP SE, SAS Institute Inc, Sisense, Tableau Software Inc, Teradata Corporation and Tibco Software IncUse Cases of Predictive Analytics in Manufacturing:
- Predictive Maintenance: Method of predicting maintenance requirements in machines on a factory floor using “Predictive Analytics” is called as Predictive Maintenance. Analytics is used on machine operational data by analysing maintenance patterns allowing operators to predict when maintenance will be required on any given unit, hence maintenance can be proactively planned proved to be less costly.
- Demand Forecasting: Usage of advanced predictive analytics, such as machine learning and explainable AI, for more accurate and dynamic projections of demand, planning production capacity and supply flows accordingly.
- Production Capacity Planning and Management: Predictive analytics is used to optimize production scheduling. Data from number of orders, available raw materials, inventory levels, demand
- Inventory Optimization: Using predictive analytics for deriving accurate forecasts from production, suppliers and customers helps in optimizing manufacturing inventory levels.
- Workforce Scheduling: Predictive analytics removes much of the time-intensive labour through automation. Combining a precise demand forecast with the number of resources on hand helps a planner develop a more profitable, optimized schedule.
- Customer Satisfaction: Predictive analytics is being used by manufacturing to derive a 360-degree view of customers that encompass satisfaction index, customer lifetime value, churn prediction and propensity to buy.
- Predictive Quality: Root-cause analysis over past data and deep level of “Predictive Analytics” helps in identifying the reasons behind the substandard batches. Which in turns helps to make prior adjustments to save the under-production batch.
Case Studies of Predictive Analytics in Manufacturing:
IBM Corporation
Case Study: IBM’s predictive maintenance solution solved IEC’s reliability issues
IBM helped Israel Electric Corporation (IEC), the primary electricity provider in Israel by offering predictive analytics solutions to keep sustainable and reliable generation of electricity specially during peak demands. The solution used the patterns of circumstances surrounding past power outages, and helped predict and prevent future failures.
Business Outcome:
- Reduce costs by up to 20% by avoiding the need to restart turbines after an outage.
- Saved approximately $75,000 in fuel costs per turbine by identifying inefficient fuel usage.
- Increased the efficiency of maintenance schedules, costs and resources, resulting in fewer outages and higher customer satisfaction.
- Provides early warning of certain types of failure up to 30 hours before they occur, instead of 30 minutes.
Microsoft Corporation
Case Study: Microsoft’s predictive analytics solution minimized the cost and disruption of maintenance for Rolls-Royce
Microsoft through its Cortana Intelligence Suite for Predictive Analytics solved Rolls-Royce problem of operational anomalies of its 13,000 engines used in commercial aircraft worldwide. The solution used wider sets of operating data and using machine learning and analytics to spot subtle correlations, which optimized various maintenance models and provide insight that improved maintenance plan and help reduce disruption for their customers.
Business Outcome:
- Provide the most accurate overview of the health of its aircraft engines.
- Even a 1% saving on fuel costs can save an airline $250,000 per aircraft per year.
SAS
Case Study: SAS’s demand forecasting solution is making the right product available to Asian Paint’s customers
Asian Paint’s struggle to consistently and accurately plan sales demand was solved by SAS solution of “Predictive Analytics”. The demand forecasting solution factored in the deviations from business functions like production, inventory, supply, distribution planning and accurately forecasted future pricing, promotions, events and stock out data.
Business Outcome:
- Improvement of forecast accuracy over the incumbent process.
- Flexibility in leveraging impact of seasonality, pricing, promotion & other key events.
SAP
Case Study: SAP helped Daikin to improve customer satisfaction by predicting quality risks
Daikin’s motive to drive customer excellence through proactively detecting air-conditioning installation quality risks was achieved through SAP’s Predictive Analytics solution. The solution optimized and renovated quality assurance (QA) process for AC units using Machine Learning and image recognition.
Business Outcome:
- Accurate prediction of potential installation issues with air-conditioning (AC) systems.
- Prioritized installation by quality risk, enabling Daikin engineers to maximize QA efficiency with minimal effort and increase customer satisfaction
Mu-Sigma
Case Study: Mu-Sigma helped a leading consumer goods manufacturer in reducing raw material inventory levels
A manufacturing excellence group was aiming to reduce the overall inventory levels include raw material, packaging material and finished good inventories in distribution centers. Mu-Sigma developed predictive analytics capability within factories by creating a detailed continuous improvement framework for inventory reduction.
Business Outcome:
- Enabled lower inventory at factory levels
- Ability to scale the solution framework globally
QlikView
Case Study: QlikView offered “single source of truth” to drive strategic decision making
Ramkrishna Forgings a manufacturing company was facing challenges related to access of information through multiple sources and make intuitive decisions. QlikView integrated its analytics platform to provide “single source of truth” which helped in strategic decision making.
Business Outcome:
- Improved data accuracy by up to 40%
- Reduced IT support for business intelligence by up to 90%
HPE
Case Study: HPE solve HIROTEC problem of unplanned downtime
HIROTEC Group, one of the largest private automotive manufacturing company wanted to implement advanced technologies to tackle the unplanned downtime in its manufacturing facilities. Hewlett Packard Enterprise (HPE)’s predictive analytics solution captured data from eight CNC machines and performed real-time visualization of entire production facility.
Business Outcome:
- Gained real-time visibility into its business operations
- Accurately predict failures in critical systems like robotic arms
Tableau Software
Case Study: Tableau advanced analytics brought operational efficiencies for Tesla
Elon Musk’s goal to produce 1 million Tesla cars by 2020 failed to meet its projections more than 20 times in the past 5 years. As part of ongoing effort to tame chaos and improve manufacturing efficiency, Tesla turned to Tableau for advanced data analytics capabilities for root-cause investigation, quality defect tracking etc.
Business Outcome:
- Tesla could easily trace back the production defects and rectify the same in no time
- Improved operational efficiencies by predicting the production count and yield ratio
Tibco Software
Case Study: Brembo used TIBCO’s predictive analytics solution to build a Smart Factory
Brembo an Italian-based braking systems manufacturing company wanted to digitally transform and develop a smart factory. Tibco offered an analytical solution for manufacturing, process optimization control, purchasing, quality control, and R&D. The solution was used for cluster cooling curves, predictive maintenance, and noise analysis and testing.
Business Outcome:
- Provided 360-degree view to organizational processes thus enabled better command and control
- The advanced R&D department used the platform to develop KPIs for noise analysis and testing
RapidMiner
Case Study: A leading silicon wafer manufacturer was under increasing pressure of quality requirements with skyrocketing demand of its wafer. One such critical quality process is wafer polishing. The more precisely the manufacturer can determine how much abrasion will be necessary to achieve the desired result, the less waste of silicon will occur (from overpolishing) and the less likelihood of a defect occurring (from underpolishing). The manufacturer is using RapidMiner to build machine learning models that predict exactly how much polishing will be needed for each wafer.
Business Outcome:
- Improved product quality and product yield
- Achieved better financial results