Emerging technologies such as the Internet of Things, machine learning and artificial intelligence are having a profound impact on manufacturing worldwide. However, the industry in Australia has been comparatively slow to embrace these innovations and risks missing out on the opportunities they represent. By Scott Hubbard.
By now, you have probably heard numerous predictions for 2020, refering to billions of devices connected to the Internet of Things (IoT) and the masses of data that will be generated. There’s no doubt that the big data created by IoT devices is increasingly driving manufacturing intelligence to generate greater operational efficiencies across the assembly line. It is helping manufacturers to target specific pain points, such as improving machine productivity and maintenance, as well as to predict supply shortages and equipment maintenance needs and failures.
Already, 40% of the world’s IoT devices are now used in business and manufacturing. However, the Australian manufacturing industry has been relatively slow to adopt smart technologies, with recent reports estimating we’re sitting at about half the global rate.
So why haven’t Australian manufacturers jumped on board the IoT bandwagon?
High energy costs and strong overseas competition have created some reluctance to invest in new technologies. However, the IoT will be the golden ticket to Australia’s continued economic success and will deliver a massive injection to further boost steady manufacturing sector growth.
The role of machine learning in leveraging IoT data
Technologies such as machine learning (ML) and artificial intelligence (AI) that can help analyse the masses of IoT data being created are transforming the industry, helping manufacturers to increase their yields and better manage plant operations. In fact, according to PriceWaterhouseCoopers, manufacturers’ adoption of machine learning and analytics to improve predictive maintenance is expected to increase by 38% in the next five years.
AI and ML can help manufacturers better predict supply shortages and/or demand spikes to help them make more intelligent business decisions regarding production output in real time. By predicting demand shortages through the use of advanced ML and AI, manufacturers also have the ability to better plan maintenance activities and prepare for downtime periods of low-customer demand — which will, in turn, optimise margins. By leveraging machine learning capabilities, manufacturing equipment is projected to be 10% cheaper in annual maintenance costs, reduce downtime by 20% and reduce inspection costs by 25%.
Supply chains also reap significant benefits from the deployment of machine learning. McKinsey & Company predicts machine learning will reduce supply chain forecasting errors by 50% and reduce lost sales by 65% with better product availability.
Additionally, analytic and machine learning capabilities provide architects the opportunity to experiment with in-place analytics. Machine learning models can also be deployed on target cloud environments without architects having to copy data in separate locations for advanced analytics.
The importance of data consolidation
To transform the chaos of raw data into constructive action, manufacturing operations heads, CIOs or plant managers should be seeking a versatile, consolidated data platform, which enables sophisticated analytics and machine learning capabilities. Through data consolidation, manufacturers can effectively drive better real-time and historical data analysis and modelling, and open the door of opportunity for better automation and machine learning.
However, when data is scattered and siloed, the integration and analysis of data becomes a tedious and resource-intensive manual process. The longer the time taken to analyse data, the more it decreases in value.
Solution architects must select the right platform that can handle the scale of Industrial Internet of Things (IIoT) data, along with the capability to process the data where it is generated. This will enable plant managers to create real-time data analytics to make informed, strategic real-time decisions.
Manufacturers must also decide which platform is the best suited to them, whether it is a single, multi or hybrid cloud, on premise or at the edge; the options are extensive. Otherwise, data can be processed on the edge near the equipment. This decision will depend on the amount of data required to take accurate action and the level of urgency of converting insights into real-time action. To maximise equipment uptime, while ensuring the reduction of regular maintenance costs, manufacturers can build a connected factory solution by applying IoT methods to their equipment and developing predictive maintenance schedules with failure analysis models.
For Australian manufacturers of today to maintain an advantage in a highly competitive sector, the use of IoT data and AI and ML technologies are critical to achieving greater efficiencies and enabling better business decisions and outcomes.
Scott Hubbard is the Managing Director – Australia & New Zealand at MapR Technologies.