Image: Mixabest, used under CC BY-SA 3.0 licence
Rob Stummer, CEO Asia Pacific of ERP software company SYSPRO, argues Australian manufacturers should explore ways to develop application-specific AI and find new avenues to enhance existing processes with AI to achieve significant improvements in productivity.
The Australian Productivity Commission has blamed falling investment in research and development and a lack of innovation by local companies for the recent mediocre productivity growth.
Australia is not alone in facing a productivity slowdown as developed economies globally look to overcome the same challenge, which is undoubtedly the catalyst for the wave of technological change happening today.
There are concerning similarities between the current slump and that of 2008 when crashing industrial production and a contraction in trade flows were signs of the depth of the global financial crisis. Ominously, production and trade are again weak, and local economists do predict recessions about every decade.
So how can Australian manufacturers protect themselves and achieve significant productivity growth without large increases in capital expenditure or labour?
According to Accenture, there has been a marked decline in the ability of traditional levers of production, capital investment and labour, to propel economic growth. It says artificial intelligence (AI) is a new factor of production that has the potential to introduce new sources of growth, change how work is done and reinforce the role of people to drive growth in business.
Accenture predicts AI could double annual economic growth rates in 2035 by changing the nature of work and creating a new relationship between man and machine.
McKinsey Global Institute said digitally-enabled advancements like AI have the potential to create growth value equivalent to efficiency improvements of a staggering 15 to 20 percent in manufacturing, translating to an additional $2 trillion in value by 2020 in manufacturing and supply chain.
All the other reputable research firms agree with the huge growth trajectory of AI, with Gartner predicting the business value created by AI could reach $3.9t by 2022, and IDC forecasting worldwide spending on cognitive and AI systems to reach $77.6b by 2022.
Here are five ways that manufacturers can raise their productivity levels by using AI:
- More accurate demand forecasting – Using AI and machine learning (the engine behind AI), computers can test thousands of mathematical models of production possibilities and be more precise in their analysis while adapting to new data such as new product lines, supply chain disruptions or unexpected fluctuations in demand. Manufacturers can use historical data to predict the future. From a seasonal perspective specifically, they can predict when a customer might purchase a particular product.
- Predictive maintenance – Manufacturers are recognising that predictive maintenance is worth the investment, because it’s a sure-fire way to improve operating efficiency and therefore has an almost immediate impact on profitability.Predictive maintenance uses sensors to track the condition of equipment and analyses the data continually, enabling equipment to be serviced as and when needed instead of at scheduled service times, minimising downtime. Machines can even be set up to evaluate their own conditions, order their own replacement parts and schedule a field technician when needed. Algorithms can use big data to predict future equipment failures. McKinsey found that AI-enhanced predictive maintenance of industrial equipment can generate a 10 percent reduction in annual maintenance costs, up to a 20 percent downtime reduction and a 25 percent reduction in inspection costs.
- Anomaly Detection – This is where an AI engine can ‘listen’ to operator activity and is typically used to capture fraudulent transactions or even incorrect keystrokes. AI is able to do this by sifting through a database, detecting the norm for particular users, and comparing this to all transactions. For example, posting of sales orders, requisitions and purchase orders. If any anomalies are found, the relevant people are immediately alerted.
- Optimising manufacturing processes – By the end of 2019, it is expected there will be a huge number of machines using machine learning algorithms capable of autonomously improving the efficiency of manufacturing processes. AI systems will monitor quantities used, cycle times, temperatures, lead times, errors and down time to optimise production runs.Initially, AI will be in “operator assist” mode where it will run in the background and suggest answers to the operator. AI systems will use the operators’ final decisions to learn how the human mind performs so it can be deployed in an “operator replace” mode. The next step will be for AI to enable the transformation of data into intelligence in a vendor-agnostic environment where all machines speak the same language, increasing production efficiency from machine to machine across the shop floor. Advanced analysis of granular data on machining processes, generated in real time, will be fundamental to identifying and addressing the underlying causes of process inefficiencies and problems with quality, faster and more effectively. In addition, forecasting processes that draw heavily on big data already can drastically reduce inventories and improve service levels.
- Automated materials procurement – Analytics combined with machine learning will record and critique everything, including the beginning stages of quoting and establishing the supply chain. McKinsey predicts machine learning will reduce supply chain forecasting errors by 50 percent and reduce costs related to transport/warehousing and supply chain administration by five percent to 10 percent and 25 percent to 40 percent respectively.
Summary
There’s no doubt that manufacturing is one of the sectors already leading the way in the use of AI, but we are still in the early adopter stage according to Gartner, with much of the work still in the research and development labs. From significant cuts in unplanned downtime to better designed products, manufacturers are using this technology to apply AI-generated analytics to data to increase efficiency and productivity, improve product quality and enhance the safety of employees.
Australian manufacturers should be exploring ways to develop application-specific AI and finding new avenues to enhance existing processes with AI to achieve significant improvements in productivity.
