Machine learning can be used to reduce risks related to environmental, health, employee safety, and quality (EHSQ) issues; in general, though, it must be used in conjunction with predictive analysis if it is to be most effective. How can organizations start thinking about the best ways to harness machine learning as the technology advances? Begin by looking at the business issues that are most critical and the challenges that most affect an organization, and plan for technology that will address those issues first, suggests David Vuong, product manager of analytics at Medgate, an EHSQ software solutions company.
Machine learning is using artificial intelligence to allow machines to “learn” without being explicitly programmed. Examples of this in everyday life include Netflix (when suggested movies that you might like pop up based on your viewing history) and Google (when you begin typing a search term and it finishes your sentence for you). But machine learning is quantitative and is not necessarily predictive analysis in its fullest form, says Vuong.
A predictive analytics program that harnesses machine learning capabilities, Vuong says, can use anonymous industry data and discover where risks related to EHSQ exist. EHSQ executives can then make decisions on how to reduce those risks, based on the available data. Within five years, companies that have systems connected via IoT will be able to benchmark themselves against others in their peer group, see correlations in their data that they may have not been aware of, and receive predictive and prescriptive insights that will help them improve their EHSQ programs, he says.
But in order to fully reap the rewards of machine learning and analytics in EHSQ, companies need to consider the best way to reach those goals early on. Companies should look at considerations such as security issues, the company’s most pressing business problems, and, of course, the budget, said Stuart Payne of Gibson Energy during the recent Big Data, IoT and Machine Learning in Oil & Gas Conference.
But perhaps the biggest considerations when looking at the future of machine learning and analytics, in terms of how a company will be able to leverage those abilities in coming years, is to look at the organization’s existing technologies and how they will interact with a new analytics program. Many companies are currently using a siloed approach to IoT, with various management systems working independently and not tied into each other. This could cause problems in the long run because it is difficult to pull data from the various systems together in order to analyze and use the data for EHSQ improvements. Carefully consider all new technology to ensure it can talk to other systems, thereby making it “future-proof” as well as scalable, Vuong suggests.
As machine-augmented decision making evolves and as organizations begin to take advantage of the technology, an EHSQ supervisor will be able to “see” and “hear” much more than before, leading to a more effective EHSQ team that can do more and miss less. “It will be a game changer,” Vuong says.
David Vuong is the Product Manager of Analytics at Medgate, where he oversees the product development roadmap for Medgate’s Analytics solution. He joined the Product Management team as the Product Manager of Business Intelligence in March 2015, where he developed a long-term plan to elevate Medgate’s Business Intelligence suite to world-class levels. Prior to Medgate, David was in the Business Intelligence industry for over five years where he led new and established best practices in data visualization and design. He has worked with clients in industries such as mining, telecommunications, and logistics.