Everything is becoming increasingly connected within the manufacturing and automotive domains with volumes and volumes of data being collected each day. While data may not have been well leveraged in the past due to the lack of adequate analytics tools and services, the availability of new methods now enables us to analyse and filter this flood of information to obtain crucial data points.
In turn, these insights help design manufacturing activities more efficiently and increase quality through targeted analysis of root causes.
Driving Operational Efficiency
As it is, big data is already being mined to improve operational efficiency, and the ability to make informed decisions based on the latest information is rapidly becoming the norm. In fact, in PwC’s 18th Annual Global CEO Survey, “80 percent of CEOs felt data mining and analysis were the most strategically important topics to focus on”. The reasons are obvious: over half of the CEOs felt digital technologies (including data analysis) created high value in areas such as operational efficiency and customer experience.
In the automotive industry, successfully managing and improving a fleet’s performance requires multiple levels of data across a vehicle’s lifecycle, and tying that data together is key to getting comprehensive insight.
For example, odometer readings are a key metric used to drive many fleet decisions. This data is currently captured through a variety of processes and the data captured is subject to multiple errors leading to poor decisions or wasted time reconciling incorrect information with the right information. Once this information is available automatically, fleet managers can spend more time on quality decision making, rather than data quality issues.
Data offers incremental opportunities for improvement if it is integrated and actionable and the push for the use of big data analytics in mobility application will thus continue to grow as more data is integrated across the lifecycle of the vehicle.
In a separate survey, Harvard Business Review found the top benefit of data mining to be decreasing expenses followed by finding new avenues for innovation. To illustrate the power of crunching customer data: computer algorithms can slice and dice everything from a customer’s frequent orders, buying history, credit scores and even gender. By carefully mining this information, analytics software can help identify patterns in customer behaviour that can increase sales and reduce customer turnover.
Predictive analytics, which uses many techniques from data mining, machine learning, and artificial intelligence to analyse current data to make predictions about the future, can show ways to speed up the supply chain and reduce the potential for delays and bring down product failure rates. When fewer products fail, manufacturers can save money on supplies, materials, and warranty costs.
In manufacturing, companies can use predictive analytics to forecast demand. Using historical data, firms can identify consumer trends and prepare for changes in volume. Forecasting demand might prevent wasted time and resources and help the firm increase customer satisfaction rates.
Implementation Into The Workplace
Companies need to develop a plan that brings together data, analytics, tools and people to create business value and an environment where data driven projects and thrive.
This begins with sourcing for data creatively. Companies can impel a more comprehensive look at information sources by being specific about business problems they want to solve or opportunities they hope to exploit. The IT department also needs to be well equipped. Indeed, IT departments may not have the necessary infrastructure to furnish continuous flows of information for real time decisions, and getting to this stage could take years. However, business leaders can address short-term big data needs by working with CIOs to prioritise requirements.
This means quickly identifying and connecting the most important data for use in analytics, followed by a clean-up operation to synchronise and merge overlapping data and then to work around missing information. New cloud-based technologies and services may also offer ways to quickly and flexibly avail of the necessary analytics tools and storage requirements meeting big data demands cost-effectively.
Nevertheless, it appears that production managers remain reluctant putting their data on the cloud due to concerns over data security and intellectual property (IP) protection. However, by maintaining certain best practices, businesses can leverage on the benefits of cloud adoption whilst minimising these drawbacks.
Small And Achievable Goals: It is generally advisable to start with a very specific and small goal with existing data sources, such as condition monitoring, and then scale up to predictive maintenance. Over time, you can polish the analytics models until you gain the desired insights. And then you can gradually increase the scope with additional models and layers of new data sources. This will not only fine-tune your analytical insights before you scale but will also help you save many resources.
Data Veracity: Given the large volumes of data, e.g. machine data, and the velocity at which it travels through multiple systems, it is very important to ensure the data is well-formed and correct. If not cleaned and audited, this may have a butterfly effect resulting in erroneous decisions and bad business.
Compatibility: The key stakeholders should verify that their environment with respect to devices and applications are compatible to the protocols, APIs and the services provided by the cloud based analytics services.
Security And Privacy: Companies’ stakeholders (IT) should ensure that their chosen cloud-based analytics services have obtained the right certifications and have implemented the basic security measures such as:
- Data encryption & masking
- Data retention policies
- Audit trails
- Advanced authentication mechanisms
The IT department should also de-identify all the personally identifiable information (PII) from the data before it leaves their data centre.
Together those approaches establish an infrastructure that propels innovation by facilitating collaboration, rapid analysis, and experimentation.
Impact Of Big Data And Analytics On Manufacturing
In manufacturing, companies can use predictive
analytics to forecast demand.
In their recent “How Big Data Can Improve Manufacturing” report, McKinsey & Company described how big data and analytics have streamlined processes in manufacturing. The first would be accelerating the Industry 4.0 vision through integrating IT, manufacturing and operational systems.
Big data is already being used for optimising production schedules based on supplier and customer information, machine availability and cost constraints. On the supply side, data analytics helps manufacturers view product quality and delivery accuracy in real time, making trade-offs on which suppliers receive the most time sensitive orders.
People Will Still Make The Decisions
Despite the call for Industry 4.0 to drive rapid sales growth, connected manufacturing comes with its own set of problems, one of the most prominent ones being the need for coordination and decision-making. After all, decision making, the “last mile” of any business journey, will still largely be carried out by people.
Thomas Jakob believes that big data
offers incremental opportunities for
improvement within manufacturing
Ernst and Young found that analytics consumption takes place at two levels: within the organisation, insights help decision makers understand their markets, product or service positioning and operations. Individually, analytics helps employees at all levels and locations throughout the enterprise improve their business processes. At both levels, data analytics enhances rather than replaces decision making.
In conclusion, manufacturing operations can significantly benefit from big data analytics and developing the right kind of technologies and expertise to apply them will be a critical skill for manufacturers to stay competitive.
Nevertheless, to maintain and further develop the ability to leverage such innovative technologies within organisations, decision makers will continuously need to enhance their understanding of how these technologies can help them to achieve their business goals. This can best be realised by proactively evaluating technologies for their benefits at an early stage, and by gaining experience with their application (and their impact on an organisation’s KPIs) by applying them in experimental settings (aka Proof-of-Concept) before later scaling them up for company-wide usage.
APMEN Feature, August 2017