The Hidden Applications of AI and Automation

By Tamara Wilhite posted 02-26-2020 11:35:31 AM

I attended a joint IISE student/professional chapter meeting at the University of Texas at Arlington last week. One of the presentations was about the evolving software and programming expectations on IE students. For example, they often learn Python but learn how to use it to process data as part of a core course. You can learn the programming language R in the DataCave, while SAS is still taught along with a variety of simulation software suites. They've improved the robotics lab since I was there, and knowing the basics of programming CNC machines is common.

Yet it is the hidden applications of artificial intelligence and automation that are affecting the profession, whether you're working in supply chain or engineering. The rest of this article is draw from an answer I provided on Quora.
What are some examples of robotic process automation?

I’ve worked roughly a decade in engineering IT. Most of the examples I’m going to share come from that and tech support documentation.

Product Data Management

Instead of having databases of drawings and manual tracking of product build lists and part changes, product data management systems do so automatically. If you list an alternate part for part number 1234567–1, everyone else who uses that part is informed of the alternative and asked if that’s OK to use on their product. When part 1234567–1 is flagged as obsolete, such as when the manufacturer no longer makes it, everyone who uses it knows they need to start researching alternatives. And they can pull from the list of alternatives being used on similar products.
The automation includes drawing and documentation updates. Anyone can review the manual and send a note tracked by the data management system that there is a misspelling on page 38 or the screen shot on page 92 is out of date. All of this is collected in centralized to-do lists. Document control can decide when to combine all these tasks into the next documentation release as well as assign the work to various people. When the new release comes out, everyone who needs to be informed is informed.
This type of robotic process automation reduced the number of people working in drawing control by half, at least, when it was rolled out at one of my employers. The part management and reporting I listed above reduced the number of configuration managers and data managers by a third. And most of them were paid engineering and/or manager level salaries.

Product Verification Integration

There are two main data sets for production. One is the as-designed list, the bill of material for the design from engineering. The other is the as-built list showing the pieces pulled from inventory and subassemblies manufactured in house that were used to build serial number 83. Before you ship, you have to review the as-built against the as-designed data to make sure you’ve built what the customer ordered OR the latest version of the product.
Robotic process automation now includes pulling the two reports and comparing them, flagging differences. This review used to be a manual process involving pages of print outs. Humans still decide whether to fill out a contract waiver to send to the customer or send the item to the shop for rewiring. And people still do the work as items come back from the field, get reviewed, and have various hardware and software upgrades done to them to bring them up to the latest version.

Project Management

Project management is definitely affected by robotic process automation. Create your schedule and task list in the project management system, and everyone is informed as to what they should do and when. You ideally see fewer status meetings because managers track everything via a project dashboard. Financial metrics get tracked as expenditures are made, though it is still a human’s job to alter the budget. This is used everywhere from construction projects to content creation.
One of the ripe areas for robotic process automation is legal management. Tracking the status of a case, reminding people of filing deadlines, and making sure money is collected / distributed.
We can expect to see robotic process automation become commonplace, reducing the number of people required to manage complex projects. And it is increasingly available through sites like Basecamp to small business owners.



10-01-2020 10:13:37 AM

Good point on what to look as potential areas to apply new ideas. For some reason even though there are some success cases around the whole area has still a surprisingly low rate of adoption.