chess champion in tournament play? And the IBM Watson cognitive system introduced in 2010 that beat the top two Jeopardy champions? Those were special purpose built systems. Now we are moving toward more of a general purpose or mainstream AI that can be utilized in many industries, and manufacturing is a prime candidate for that.
This has been enabled by a long list of advances in technology across the board. Worth mentioning are open source, cloud computing, the Internet of things, machine learning, deep learning, Big Data analytics, and high-performance chips and devices that have become more affordable. Implementation in manufacturing continues to grow and companies are beginning to use AI and machine learning to tackle some well-known challenges such as maintenance, quality, production optimization, improvement of overall equipment effectiveness (OEE) and predictability of events in plants globally while augmenting HI with AI to figure out the next best action and track execution to improve overall key performance indicators.
PlasticsToday: Will AI give us all the benefits we expect—reduced labor costs, shorter unexpected downtime and increased production? How will the role of human-AI interaction change over the next five years?
Almasarweh: As I mentioned, the platforms that enable innovation in AI are becoming widely available through IBM Watson technologies, Google AI, Microsoft AI and the thousands of developers and startup companies already investing in AI. They will move this faster than any other technology innovation in human history. Industry leaders are talking about a race between the major economic powers to achieve the most in AI. The United Arab Emirates has just established a ministry for AI to drive its use in the greater economy. So, there is little doubt as to its potential to generate tremendous value. In manufacturing, taking out costs and improving a plant’s OEE, which means a need for optimized manufacturing asset uptime, are all important. I think that HI augmented by AI systems, which continue to be trained with every new event and interaction, will help tremendously in achieving these goals. I also think that there will be more and more use cases where AI will deliver immediate value and that will vary from one industry to the other.
Some promising applications are using AI to build custom configurations of products based on consumer demand with minimal human intervention. In some cases, robots can interact with humans in a collaborative way as events happen during the manufacturing process, potentially anticipating market demand then managing supply chains across multiple geographies, weather and social events.
We live in an interesting time, and surely manufacturing is a prime area to benefit from all the advances in AI in the coming years.