Making RPA a unifying force for a vast, diversified conglomerate

To understand the importance of robotic process automation (RPA) at Flint Hills Resources (FHR), it’s important to understand the technological tangle its parent, Koch Industries, is working to unravel.

Koch has grown to become one of the world’s biggest private companies, in part, through acquisitions. And with those acquired companies came a patchwork of legacy technology that, over the years, grew to more than 50 enterprise resource planning (ERP) systems, including some that were decades old, and more than 7,000 custom applications.

The $120 billion diversified conglomerate has spent four years streamlining those systems and standardising the data that runs through them. The goal is to get better use out of information it already possesses and to enable accountants to spend more time analysing and less time importing and verifying historical data.

“Looking 100% in the rear-view mirror is a good way to drive off a cliff,” said Richard Dinkel, CPA, a Koch vice-president who serves as the company’s controller and chief risk officer.

Once finished, Koch hopes to have about 12 integrated ERP systems that would enable the corporation to complete tax returns the minute the fiscal year is over and provide outside auditors with instant access to up-to-date ledgers. Subsidiaries, meanwhile, would more easily share pricing data to ensure they’re getting the best deals from vendors. Manufacturing facilities, with the help of machine learning, would be able to better predict when a part might need maintenance or replacement. And cognitive artificial intelligence would enable the creation of automated forecasts, perhaps saving thousands of employee-hours in the process.

The transformation could take another five years, which is why RPA is a critical tool. The software plays well with a variety of user interfaces, making it a useful stopgap as Koch and its subsidiaries connect platforms and processes that haven’t yet been fully integrated. As more processes are automated, more human resources are freed up to help the integration along.

“It also allows us to experiment or prototype processes without investing a lot in new platforms or integration between platforms,” said Gary Matula, Koch’s chief information officer.

Here’s what that looks like on the ground: To aid systems integration, Koch last year acquired 50% of the software company Infor, which builds enterprise cloud software. Soon after, FHR replaced its on-premises general ledger system with Infor’s CloudSuite Financial tools. FHR is now in the process of replacing its maintenance management system with another Infor product.

Using RPA to automate legacy processes helped lower the cost of the projects and shorten implementation windows. That’s because the employee-hours gained through RPA freed up employees to focus on implementation of the new software, said Scott Bender, CPA, CGMA, director of automation and modernisation at FHR.

Once Koch’s ERP modernisation project is complete, some subsidiaries might do away with RPA. Others, such as FHR, are Making RPA a unifying force for a vast, diversified conglomerate expected to continue using RPA to streamline processes. The latter scenario will require acute management if Koch’s digital strategy is to succeed.

Streamlining data collection efforts

Data, of course, is the critical ingredient here. Koch wants data science to permeate its operations — from corporate finance to the factory floor. To that end, Koch has purchased at least two data analytics companies, bringing in more than 100 doctoral-level data scientists to help harvest the corporation’s existing data, Dinkel said.

The key, though, is ensuring the data is clean and usable. That will require standardisation. “If you don’t have consistency or standardisation or a common governance process, you’re going to get out of control in a hurry,” Dinkel said. “The more transactions you can do with digital labour, the faster you can get out of control when you don’t have humans at the wheel.”

Achieving consistency requires a delicate balance of corporate governance and subsidiary empowerment. The parent organisation provides subsidiaries with guidelines for the information it needs to connect the data dots at a higher level. In that process, there’s room for debate over what’s too much information and what’s not enough. Once corporate data collection terms are agreed upon, the subsidiaries are free to operate as they see fit.

Providing guidelines at the corporate level and autonomy at the subsidiary level allows each company to collect data and enforce security and controls in a way that makes sense to the individual business. It also creates an environment of experimentation that, when communicated, leads to efficiencies across the companies.

“It’s better to have ten individual laboratories than one,” Dinkel said. “… [When] you can learn and share the results, you’re going to get progress quicker.”

Photo by Alex Della Gatta/AP Images