Four AI use cases in industrial and manufacturing companies

01 August 2019

The use of artificial intelligence can create major benefits to industrial companies. Aime Lachapelle, an expert in data & artificial intelligence transformation, spoke with to reflect on four trending use cases emerging in sectors such as aviation, automotive, chemicals, manufacturing, logistics and high-tech manufacturing.

Having spent a decade in consulting and business, Aime Lachapelle currently is Managing Partner of Emerton Data. Over the past years, he has noted a number of sea changes across the industry, and believes it is time “to acknowledge the challenges to delivering significant impact from the use of AI in manufacturing industries.” After surveying dozens of industry experts in the industrial sector, Lachapelle found that while 85% of respondents believe they need to implement AI on their production processes, less than 30% actually have it in place.

Maintenance and robotics

Today, poor maintenance strategies can reduce a plant’s overall productive capacity by as much as 20%. Recent studies also show that unplanned downtime is costing industrial manufacturers an estimated $50 billion each year. According to Lachapelle, this begs the question, “How often should a machine be taken offline to be serviced?”

“Traditionally,” he explained, “this dilemma forced most maintenance organisations into a trade-off situation where they had to choose between maximising the useful life of a part at the risk of machine downtime or attempting to maximise uptime through early replacement of potentially good parts with time-based preventive maintenance.”

Four AI use cases in industrial and manufacturing companies

The problem is that with limited budgets, maintenance professionals must evaluate which parts they’ll need and when to procure them. If the spare is not on hand or on order when it’s needed, the downtime of an asset can be anywhere from days to weeks – or even months – while waiting for the replacement part. This typically leads to the build-up of spares inventory, which not only ties up working capital, but also increases the risk of excess and obsolescence that erodes the bottom line.

Fortunately, there is an alternative. Predictive Maintenance (PdM) is primarily designed to help determine the condition of in-service equipment, and predict when maintenance should be performed. When executed correctly, this approach can avoid costly down-time of machines compromised by neglect, instead overseeing a smooth transition of much-needed new parts and repairs. As a result, the technique promises cost savings over routine or time-based preventive maintenance, because tasks are performed only when warranted. At the same time, this technique looks to have been strengthened by new machine-learning technology, which can increase the accuracy of the predictive algorithms, leading to even better performance.

“AI industrial robotics is about combining machine learning with industrial robotics technologies into a single system, used for manufacturing production,” Lachapelle explained. “AI embedded in industrial robots allows engineers to deal with the main concerns of the current generation of robots: safety in close proximity to humans, ability to reconfigure quickly and ability to grasp delicate items.”

According to Emerton’s analysis, recently recognised as a leading strategic and digital consultancy to the industrials sector, today’s main focus areas for AI industrial robotics are threefold. First, it must communicate to develop collaborative-robotics, improving the seamless interaction between workers and machines. Better communication solutions can reduce the cost and duration of worker training, especially since turnover is costly in the industry. Then, AI industrial robotics must focus on navigation and picking capabilities, to allow robots to optimise their movement within the factory and navigate in non-standard environments. Finally, the third focus area is learning, in order to enhance robots’ programming via physical/video demonstration, to facilitate trial, error, and ability to learn in a group, as parallel computers.

Lachapelle elaborated, “Overall, AI embedded in industrial robots is expected to directly improve uptime, productivity, and safety. Another consequence is to reduce menial labour, by limiting repetitive tasks. Here industrials could greatly benefit from emulating winning models in the consumer space. Autonomous cars and voice assistants like Amazon Alexa are examples of how AI can unlock productivity, engagement, and collaboration with hardware, and we believe this can be duplicated in many manufacturing use cases.”

“85% of the companies surveyed state they aim at implementing AI in their production processes. However, less than 30% actually have an AI development plan. The challenges to delivering significant AI should not be forgotten.”
– Aime Lachapelle, Emerton Data

Process optimisation and vision control

Pointing to another key trend, Lachapelle noted, “Production process optimisation empowers manufacturers with 100% visibility in their plants, lines, machines, products, and process data, improving productivity throughput time and quality. Production Process Optimisation is always cited by the top industry leaders interviewed as the top 1 or 2 priority for a future use case implementation, with quality improvement as the first step. This is a master use case, expected to directly impact the heart of the factory.”

He stated there are two main sub-use cases that are expected to have a huge impact in manufacturing industries. Firstly, production settings optimisation is important because processes are often complex, made-up of multiple steps and parameters. In order to ensure proper industrialisation and return on investment, then, top players in the field are building AI algorithms on top of digital twin platforms. The digital twin platform is the only way to ensure that these solutions provide contextualised recommendations and enhance investigation environments to drive adoption. Sight Machine claims that their platform helped industrial manufacturers reduce scrap costs by 30% within 3 weeks.

The other high-outcome application is production planning and scheduling. Complex processes must be tackled in regards to the availability of raw materials, production capacity and demand. AI helps find the best optimisation strategy given machine data, supply & demand data, and recommends an optimal scheduling and maintenance plan. It positively impacts production, maintenance costs, scrap rate and customer service.

Remarking on the two sub-uses, Lachapelle said, “In order to make process optimisation successful, companies must also integrate humans either quantitatively by turning human actions into data using computer vision, for example, or qualitatively by understanding how operators interact with the plant. At the end of the day, people are still governing processes, so process optimisation products must be customer-centric and perceived by operators as tools that facilitate their work.”

Finally, he pointed to vision control as a booming use case that combines vision hardware and sensors on the production line with computer vision algorithm tools. The main objectives are often to automate the visual identification of non-quality and to find their root causes in order to prevent the production of low-quality items in the future. The manual inspection of products can sometimes be completely automated at every stage of the production line.

Despite this, analysis from Emerton Data shows that less than 30% of industrials actually have an AI development plan for their factory – even though over 85% believe they need to implement AI in their production processes.

Lachapelle concluded, “In order to exploit the full potential of this technology, vision control products should provide a collaborative platform that allows engineering teams to discover issues easily, dig deeper into failures, and implement corrective actions immediately. One of the key success factors of these platforms is ergonomic human-machine interaction, with a seamless user experience for data generation and data labelling. In order to reach acceptable levels of performance, deep learning algorithms need to learn a huge number of images of quality, non-conformities, or failures.”