(83b) How Virtual Advisors Create Practical Employee Efficiency Boosts in the Chemical Engineering Workplace

The use of machine learning to boost human efficiency has become commonplace in many industries, some of which include logistics, law, medicine, and finance. Virtual advisors, functioning in ways similar to the popular consumer products “Alexa” and “Google Home” but aimed at helping knowledge workers make better decisions faster, are at the crest of this wave, and already organizations which have failed to adopt such tools are being left behind. Going into only cursory data science depth, this presentation will explain why chemical engineering is an ideal field for custom-built virtual advisors and why chemical research and processing environments have enormous value to capture from effective management of this technology. In essence, a properly functioning virtual advisor can help the chemical engineer to find responses to complex questions in real time by referencing historical documents and understanding the relevance of the contained text and chemistry.

Major points include why dimensional reduction and scaled matrix applications of basic sorting and predicting algorithms (k-means, linear regression) work particularly well on natural language text concerning chemistry and chemical engineering, what human factors in the chemical research and processing industries are creating increased demand for efficiencies these algorithms can create in the form of virtual advisors, and best practices for management to find and employ virtual advisors to ensure employee efficiency and satisfaction.

In addition, two chemical engineering case studies will be explored, one from a research environment and one from an industrial processing environment. The first case study will show lessons learned from implementation of a virtual advisor to boost productivity of research chemical engineers and design engineers, with particular emphasis on how superior historical data reference uncovered opportunities for improvements. The second case study will show lessons learned from implementation of a virtual advisor to help process engineers resolve urgent process issues. This study will have particular emphasis on measuring how the advisor reduced the under-informed decision making that was prolonging expensive process downtime. Finally, the impact on team dynamics in both case studies will be explored.