To be able to predict deficiencies on a safety program will enable safety professionals to address those deficiencies before an injury happens
– Sergio Andrade, CSP, CHMM, MsP, CRSO | Safety Manager at Southern Star
Predictive analytics employs multiple statistical techniques including machine learning, predictive modeling, and data mining to analyze data to make predictions about future events. This approach holds great promise in the field of EHSQ, with the potential to greatly reduce the risk of employee safety and health issues and enhance the protection of the environment. Some consider it the “holy grail” of EHSQ technology innovation.
Medgate’s Analytics Program was launched in late 2016 as a multi-phased strategic initiative aimed at harnessing the power of over 6.5 million employee records and millions more data points, securely managed by Medgate on behalf of its industry leading customers.
With an initial focus on safety, Medgate has formed a Client Advisory Committee of organizations committed to improving EHSQ program outcomes, including Tri-State Generation and Transmission Association, Whiting Petroleum, Southern Star Central Gas Pipeline, Westmoreland Coal Company, Shaw Industries Group, Inc., Toyota Engineering & Manufacturing America (TEMA), and Seattle City Light.
Why Safety? Why Medgate?
Over 2.9 million non-fatal workplace injuries were reported by private employers in 2015. Each lost time injury costs approximately $100,000; employers pay over $1 billion each week for direct workers’ compensations costs. Safety programs have evolved to address these challenges in ways such as employee portals and mobile apps to make it easier for safety professionals to analyze and improve their programs. However, many safety professionals remain frustrated because while they have access to the data, they don’t have the time or know-how to analyze it. Early analytics programs have been stymied by the lack of data science resources, inconsistencies in aggregated data, or sparseness of data about rare events like injuries or fatalities.
With over 1.5 million safety events, Medgate’s Analytics Program takes a data science discipline to deliver new insights into EHSQ data across a very large and highly consistent data set, greatly enhancing the effectiveness of analytics. The goal is to provide the right insight to the right person at the right time to enable proactive interventions at a program or individual level.
“To be able to predict deficiencies on a safety program will enable safety professionals to address those deficiencies before an injury happens,” said Sergio Andrade, CSP, CHMM, MsP, CRSO, Safety Manager at Southern Star. “This is the ultimate goal for any Safety Professional.”
“Medgate’s push into the predictive analytics space tightly aligns with desire of safety managers to better leverage the increasing amounts of data they are able to gather. The growing Advisory Committee will be a critical factor in deriving maximum value and utility from these insights, and will only serve to strengthen Medgate’s offering overall. The influx of EHS data points available to safety leaders is creating an environment wherein there are countless insights available, but a lack of resources to uncover or act upon them. Medgate’s push into predictive analytics is a great step towards correcting this issue for their clients by allowing them to optimize safety performance long before an issue arises.” – Trevor Bronson, Industry Analyst at Verdantix.
Backed by Applied Analytics Experts
Medgate investor Georgian Partners is a contributor to the Program. Georgian Partners brings deep expertise in applied analytics in multiple industries and works with portfolio companies to accelerate their programs, such as Kinnser Software who pioneered predictive analytics to reduce re-hospitalization in the home care industry.
“Our thesis for investing in Medgate was based on the great potential we see for applied analytics to transform the EHSQ industry, along with their deep EHSQ domain expertise and the trust that their tier one client base puts in them to manage their most sensitive data. Analytics and machine learning hold so much promise for EHSQ, but having the right combination of rich and consistent data, with data science skills, and industry expertise is essential for success.” – Parinaz Sobhani, Director of Machine Learning at Georgian Partners.
The first phases of the Analytics Program are already generating encouraging results according to Medgate Analytics Product Manager David Vuong, “Our preliminary models are already showing substantially higher accuracy in identifying high risk scenarios versus the baseline. While our initial focus is on safety incidents, as we look forward, we intend to provide recommended actions at the strategic company level, the operational site level, and ultimately and the employee level across EHSQ.”
Our preliminary models are already showing substantially higher accuracy in identifying high risk scenarios versus the baseline.
– David Vuong | Medgate Analytics Product Manager