Which AI Methods to Use to Solve Operations Management Problems?

Artificial Intelligence (AI) algorithms can be used to perform much of the “Intelligent Grunt Work” tasks required for industrial operations management, including:

• Automated Data Capture and Analysis
• Decision Support – Advisory Information Presentation
• Automated Planning and Scheduling
• Automated Updating of Systems
• Generation of Email Alerts and Exception Reports
• Relieving People of Tedious Paperwork, Spreadsheets, and Manual Data Entry.

These can be broadly classified into two groups:

1. Intelligent Agent systems, which run autonomously to perform routine tasks on behalf of people. These typically use methods such as  rules, decision trees and self-adaptive predictors.
2. Decision Support systems, which may range from business intelligence systems which present data in easy to understand formats to chat-bots, which enable managers to ask natural language questions about the status of their operations.

 Today, over 98% of all manufacturing plants in the USA are make-to-order plants, producing semi-custom individual products or short runs with delivery times of a few days, due to the "Amazon" effect. Here, the choice of algorithms is largely driven by the need for real-time time operational data without which it is impossible for chat-bots to deliver real-time advice or intelligent agents to make timely decisions.

Fortunately, most of the algorithms used in intelligent agents and for automated data collection can be run on small inexpensive IOT (Internet of Things) computers, located in each plant or warehouse, and most business intelligence software can be run on desktop PCs.  The exception are regenerative AI based systems that require the use of large shared super computers to run their algorithms, which is expensive and can be very slow in responding.

For this reason, smart operations management systems in use today do not use regenerative AI algorithms except, possibly for performing post-facto analysis of large datasets collected by real-time operations tracking and management systems .

Instead operations tracking and management systems can be quickly and affordably implemented using systems like WIPtracker, possibly in conjunction with a system like SmartOps24x7 for integrating decision support over multiple plants and warehouses.

Please click here for more information about Intelligent Agent Systems.