It is almost a cliché today to state that modern industrial concerns are facing unparalleled competitive pressures and a cost/price squeeze, both within and across industrial arenas. Yet this is the reality that all senior corporate managers face daily, and in fact have faced since the 1980’s.
Responses have ranged from qualitative process tweaks (“Re-engineering”) to more quantitative responses (Design of Experiments developed in the 1930’s, Statistical Process Control developed by Juran and Demings in the 1950’s). Yet all these approaches are reactive in nature, need expertise to design, develop, and implement, and don’t take full advantage of modern 21st century methodologies.
Existing Market Scenario of Manufacturing Industry
We all know that manufacturing productivity is a very complex non-linear and highly interactive combination of such parameters as yield, energy/material efficiency, quality, cost, and throughput, among many others. The difficulty of balancing these factors is captured by the old process motto: “Better, faster, cheaper – pick two out of three…” But – what if management wants three out of three? Currently, there is no comprehensive model to dynamically correlate these factors into an overarching “Productivity Excellence Index” (PXi), as it were, that properly interpreted can gain insights into productivity performance and improvement.
The core of modern manufacturing is productivity excellence and its five contributing factors outlined above: yield, energy/material efficiency, quality, cost, and throughput. The ability to quantify and cross-correlate these factors in almost real time allows a company to improve its productivity and squeeze more operating profits from sales. The ability to see emerging trouble spots through analyses of latent patterns allows early interventions before the issues become dehabilitating.
A survey of current business literature indicates that industry executives understand that “AI” and “Machine Learning” have the potential, properly harnessed, to vastly improve their operational performance and profitability. But – to these executives, these concepts are just terminology that are frequently tossed around without any real corresponding comprehension as to what they really represent. Part of the underlying problem is that business concerns do not have the in-house expertise needed to exploit this technology and little idea of where to find it.
The Quantellix ML solution:
Quantellix ML has developed algorithmic expertise and its associated technology to be the company that can apply AI/Machine Learning to corporate manufacturing optimization. Our Corporate Mission is to partner with manufacturing enterprises to improve their productivity through the application of machine learning, process scoring, and benchmarking. And our clients do not need to do anything out of the ordinary during normal process operations. Our algorithms can use data from plant log sheets, already generated during standard plant operations, to reveal the deeper truths controlling plant efficiency.
We offer manufacturers access to cutting edge AI and ML technology via an easy to implement Software as a Service (SaaS) outsourced business model. Quantellix uses a proprietary formula coupled to a customized machine learning algorithm. Our algorithms dynamically measure and quantify such parameters as yield, efficiency, quality, costs, and throughput to create the PXi measure. The PXi can be used to monitor and contrast the health of the client’s manufacturing operations, and concurrently offer insights so to how the client can improve performance.
The PXi identifies and automatically optimizes the critical features of the manufacturing operation. Based on these parameters, data-based recommendations for process improvements are provided to senior management. This PXi scoring methodology allows Quantellix to quantitatively benchmark manufacturing efficiency across industry/verticals. Additionally, our clients do not need to have (or recruit) specialized AI/ML expertise – an important savings. Each of our clients is supported by an expert Quantellix ML account team, whose membership contains the expertise to make the optimization program a success. And for added simplicity, data input is conducted through standard Excel .csv files.
Our algorithms allow clients to improve their productivity and efficiency by:
1 – Developing a comprehensive score that measures the effectiveness of manufacturing operations via five productivity factors.
2 – Deriving the plus and minuses characterizing each factor
3 – Identifying the most relevant metrics
4 – Outlining insights that far outweigh standard process control data, to enable superior optimization
5 – Seamless integration with Supervisory Control and Data Acquisition (SCADA) systems
Quantellix fully intends to expand its capabilities. Our proprietary algorithms and industry benchmarks have numerous applications as we expand our business model:
1 – Logistics and supply chain optimization
2 – Expansion of AI/ML into Marketing and Sales automation
3 – HR attrition studies and optimization
4 – Optimization of retail conversion rates and average sale size.
Quantellix is ready to take your organization into the 21st century and give your organization the tools it needs to succeed in today’s hypercompetitive environment.
About the Authors
Dr. Srinivas Kilambi (Founder and CEO of DXi, LLC) has a doctorate degree in Chemical Engineering from the University of Tennessee-Knoxville. His areas of expertise include Artificial Intelligence, Machine Learning, Digital User Experience, Green Chemicals, Bio-Refineries, nanotechnology etc. He is also helping Clarkson University to build courses in the practical uses of AI and machine learning where he serves as an adjunct faculty. For additional information, contact: firstname.lastname@example.org
Dr. Joel Shertok is a practicing PhD chemical engineer with a specialization in commercial process scale-up. He feels that the combination of AI and Machine Learning with the process industries will be a powerful combination of technologies for the 21st Century.