“A few decades ago,” writes Motti Nisani, CEO of emaze, “artificial intelligence was just an exciting topic among engineers and developers. In recent years, machine learning has emerged as the ideal outgrowth of big data, breathing new life into concepts such as artificial intelligence.” The business imperative to collect and analyze data is only becoming more acute with the emergence of the Internet of Things (IoT). The IoT is going to generate oceans of data in unprecedented amounts. Machine learning and big data have a symbiotic relationship. In the executive overview of a study on Machine Learning and the Industrial Internet of Things, David White (@) and his colleagues at ARC Advisory Group explain, “Machine learning has been around for decades. However, after a flurry of enthusiasm the technology languished. More than anything else, the lack of training data in many application areas was a fundamental problem. Successful machine learning depends on data, lots of it, and the more the better.” They add, “We are poised for an era in which vast amounts of data will be generated by the Industrial Internet of Things (IIoT). That, in turn, will usher in a raft of new applications. Machine learning applications that previously were not possible — or cost effective — will now become more appealing, and hence more common.”
Commenting on the ARC Advisory Group report, Steve Banker (@) notes that the report discusses “how machine learning is helping support more efficient supply chains. Machine learning is being discussed more often in supply chain contexts because of the buzz created by the Internet of Things (IoT).” Although the widespread application of machine learning remains in its infancy, machine learning has been around for quite a while. Bernard Marr (@) notes, “Machine learning started as far back as the 1950s, when computer scientists figured out how to teach a computer to play checkers. From there, as computational power has increased, so has the complexity of the patterns a computer can recognize, and therefore the predictions it can make and problems it can solve.” The ability to solve problems is exactly why machine learning, artificial intelligence, and cognitive computing are generating so much discussion. The ARC Advisory Group report explains:
“Machine learning applications are self-modifying, highly automated, and embedded. That is, machine learning algorithms are designed to continuously adapt and improve their performance with minimal human intervention. Machine learning algorithms are also embedded within a process or workflow. That is, they become seamlessly integrated into the process to the point where they are invisible to the user or operator. Machine learning algorithms are in their element solving problems that are too difficult or complicated for human programmers to code.”
So what kinds of problems can machine learning tackle? Below are few examples.
Banker notes that one of the “use cases” for machine learning discussed in the ARC Advisory Group report deals with equipment maintenance. “In one of the more interesting use cases,” he writes, “machine learning techniques were applied against multiple data streams collected from an oil rig during operation. Deviations in the pressure readings for a water injection pump indicated that a seal in the pump would fail within days. This early warning enabled the seal to be replaced before it failed, avoiding the associated downtime. The owner-operator estimated that this data-driven predictive approach avoided an estimated $7.5m in losses from unplanned downtime.” Failures can often lead to damage. The ability to know when a piece of equipment is facing imminent failure can, therefore, save more than just downtime.
Cyber security has risen to the top of most companies risk management priorities. Danny Palmer (@) reports how the publishing house William Hill uses machine learning to fend off cyber attacks. Finbarr Joy, chief technology officer at William Hill, told Palmer, that best way to counter cyber attacks is by employing machine learning. Joy noted that “the use of machine learning and other advanced algorithmic techniques can much more quickly spot the abnormal behaviour associated with intrusions than a human can.”
Because systems that employ machine learning get better over time, their ability to make routine decisions also get better over time as does their ability to recognize anomalies requiring human intervention. By automating routine decisions, humans are freed to concentrate on other, more important, activities. This is especially useful in supply chain activities (like order fulfillment and risk assessment). A complementary capability to automated decision making is the provision of actionable insights to decision makers. In the field of medicine, for example, Marr reports, “It takes four highly trained medical pathologists to review a breast cancer scan, decide what they’re seeing, and then make a decision about a diagnosis. Now, an algorithm has been written that can detect the cancer more accurately than the best pathologists, freeing the doctors up to make the treatment decisions more quickly and accurately.”
On the consumer facing side, machine learning can also be used to make recommendations like those Amazon provides when consumers are looking to make a purchase. Machine learning helps manufacturers and retailers understand their customers better so that the offers consumers receive are more in line with their personal preferences and lifestyles. Nisani explains:
“In understanding what you are saying, machine learning will become more capable of comprehending who you are, including making deep assessments about your personality. The more we use digital devices, the more information is captured and the more we learn about ourselves. On the user end, the most immediate change will be more intensely personalized search results. The increase in knowledge about users will also reduce the number of unwanted and irrelevant ads that users are forced to see. In 2016, the digital world will become much more attuned to your personal world. Your computer will begin to create an assessment of who your best friends are, what your food preferences entail, and even what your mood may be.”
Nisani believes that the coming year will see even greater leaps in the advancement of machine learning. “These leaps,” he writes, “are likely to be made on three main fronts: natural language processing, personalization, and security.” Machine learning systems are not perfect; but, as Marr notes, “While the technology still isn’t perfect in many cases, the very concept of machine learning — that machines can continuously and tirelessly improve, they will get better.”
 Motti Nisani, “2015’s big leap into machine learning,” Geektime, 27 December 2015.
 Steve Banker, “What is Machine Learning? How will this Technology Change Supply Chain Solutions?” Logistics Viewpoints, 4 January 2016.
 Bernard Marr, “5 Ways Machine Learning Is Reshaping Our World,” Forbes, 22 October 2015.
 Danny Palmer, “William Hill uses machine learning to fight off ‘constant’ cyber attacks,” Computing, 5 January 2016.