Robotic process automation helps with data extraction
In the government use of robotic process automation (RPA), the Defence Logistics Agency (DLA) has been a pioneer to make its operations more efficient. These efforts have been accelerated by the coronavirus pandemic and the DLA is now advancing how it is using robotic process automation to become more efficient.
Frank Wood, a supervisory IT program manager at the DLA who manages the agency’s robotic process automation program says the DLA serves as “one robust supply chain from the nation’s suppliers to the war fighters”. Amid the pandemic, seeing large spikes in activity and transactions at times, the DLA has been no different than other elements of the countries supply chain, especially on in the pandemic. Robotic process automation tools helped the agency manage those spikes.
According to Wood, at present, there are 111 automations in process with DLA, out of which, 92 of them do not require human interaction. Although the number of hours of labour RPA saves the DLA is variable, currently the agency is on a path to get 122,000 hours back for its mission via robotic process automation, with another automation that is in the works potentially able to save 78,000 hours all on its own, says Wood.
He said, “Our program number will change as soon as we get that thing into production in the next several weeks”. “It’s a variable number, but we think it’s a good, sizable one”.
During the pandemic, especially robotic process automation has been proven useful. For what are known as “post-award requests,” or electronic requests for administrative action or information on a contract award, the DLA’s acquisitions team currently receives hundreds of notifications.
Those include supplier notes, comments and other information that needs to be reported to supply chain leaders for review and then provided to contract support personnel.
“Gathering all of that stuff manually in the volumes that you see under a pandemic condition, you can’t get there from here, but the bots actually enable that,” Wood says. “It’s a bot doing that sort of workflow that helped close the gap and ensure delivery in spite of the spikes”.
The DLA started its robotic process automation program in April 2018, and in the roughly three years since, the program has evolved a great deal, Wood says. Bots of robotic process automation help with a wide range of tasks.
To look for discrepancies and errors, that includes reconciliation, or processes of comparing data from multiple systems. Another is transaction processing, or updating sales orders and invoices. Yet another is compliance like responding to Freedom of Information Act requests. Also, for generating cyclical reports that need to be updated in very mandated ways, robotic process automation helps DLA.
Wood says that additionally, robotic process automation helps with data extraction: collecting the data from multiple sources, persisting the data and then taking an action that adds value to the data.
He said that RPA also helps with audit support that happens annually. “If you have information systems touching finance systems, that are very auditable stuff,” he says. “So, you need to provide evidentiary matter at the request of the auditors. When things are manual, you’re stopping your day job, which, quite frankly, is supporting our war fighters”. The bots of robotic process automation grab that evidentiary matter quickly and deliver it to auditors.
Within the DLA, robotic process automation is growing more complex. Now, the RPA program is working with process managers to identify “border processes” or leading and trailing processes so that the agency can start “chaining processes end to end and increasing the sophistication and complexity of our particular automations”, said Wood.
DLA is taking advantage of workflow improvements as this is happening. As opposed to mimicking an existing process with an RPA bot, that involves using technology to eliminate steps in processes.
From different DLA components, the DLA has also created an internal steering committee of process owners and representatives, who are working to ensure that the RPA strategy is aligned to the work the RPA program ends up doing, said Wood.
“The backlog of use cases that are requested is analysed, feasibility is checked and then prioritized by the steering committee,” he says. “That’s where the strategy meets the actual work being performed”.
To allow for DLA components to create and fund automations on their own timelines, still the DLA is also looking to create a federated model and in some cases using their own government-citizen developers, according to Wood.
He said, “That’s another evolution in terms of scaling and extending our governance, while enabling people to grow on their own particular support and command strategy”.