AI Effecting a Positive Disruption in the Manufacturing Industry

4th July 2018

Author: Anilesh Kumar, CEO, Levtech Consulting

The factories of today have started realizing the futuristic benefits of technology intervention. Manufacturers are finding new ways to grow, without compromising on product quality and excessive production lead times. Existing ERP and CRM systems have become constantly strained under the weight of growing performance markers and tightening delivery timelines. With technology advancements, Artificial intelligence (AI), Machine Learning algorithms and platforms are empowering manufacturers with refined business models, optimized and intuitive shop floor capabilities, while also helping with critical operational decision-making about what is needed, and when.

Over the last decade, terms such as automation, data mapping, machine learning and IoT have been increasingly thrown about in the manufacturing space. Intel, Funac, Bosch, Siemens, Microsoft and other companies have been investing a substantial amount of time and money in understanding and implementing cutting edge technologies either directly in-house or with third-party models. From managing labor costs, optimizing unplanned downtime, reducing product defects and operational management on the shop floor, to sales forecasting and production planning, every aspect of the manufacturing lifecycle has been observed to benefit from the capabilities of AI.

Manufacturing is slowly transitioning from a standalone traditional process to a ‘smart’ optimized one. Trendforce predicts that ‘smart manufacturing’ – the use of Industrial IoT and AI – will exceed $320 billion from the current $200 billion by the year 2020. Industrial robots too are going to be a permanent presence on factory floors, as forecasted by the International Federation of Robots.

Manufacturing Processes impacted by AI and Machine Learning

Better performance, tighter processes, increased productivity and predictive analysis will be the game changer in the manufacturing sector, with the end goal being optimizing time-to-customer performance.

• Optimised Shop Floor Performance and Productivity by IoT– Real-time knowledge of machine load level is very essential to keep track of overall production schedule, forecast production performance and prevent shop floor downtime. IoT and machine learning applications provide constant insights into production schedule performances, and shop floor productivity at every stage. German conglomerate Siemens for instance, uses specialized neural networks to monitor its manufacturing and industrial applications. Referred by the German government as ‘Industry 4.0’, the application conducts highly specialized work for the company, including reducing gas turbine emissions by a significant 10-15%.

• Early Predictions Determine Operational Accuracy – Real-time shop floor monitoring technologies, help in understanding information and creating accurate data sets to evaluate inventory, capture pricing and calculate product churn. This in turn helps determine costing behaviour across variable manufacturing scenarios, that may suddenly arise. GE’s ‘Brilliant Manufacturing Suite’, can track every aspect of the manufacturing process to detect inefficiencies linking design, manufacturing and supply chain into an intelligent predictive system.

• Robotics to Have Greater Applications in ‘Factories of Future’ – Industrial robotics are achieving greater relevancy in manufacturing for optimized operational work, and ensuring cost-effectiveness of processes. Their use in the actual product manufacturing and process work has effected a drastic reduction in industrial product errors, wastage of time and ensured expensive machinery deliver constant value. Japanese company Fanuc, a leader in the industrial robotics space began to employ deep learning and AI technology in its manufacturing robots in 2016. With a plan to use them in the ‘factories of the future’, robots will gain greater autonomy, becoming intuitive and smart enough to handle varied products quickly and manage processes efficiently.

With AI still in the evolutionary phase, but gaining significantly in impacting businesses, studies show a third of manufacturing processes becoming self-learnt and streamlined with it. Although there is no substitute for human intervention to tackle sudden challenges and unexpected demands, one can not dismiss the far reaching impact of this disruptive technology on every aspect of the manufacturing space – from shop floor to storefront.