But government agencies at all levels are getting on board quickly as they realize that as Baby Boomers retire, adoption of these systems is a relatively low-risk way to automate those workers' manual, rote tasks, leading to greater efficiencies and holding out the hope of real transformation for the way the day-to-day work of government is accomplished.
IA can be broken down into three classes. Class 1 is "robotic process automation" (RPA), a quick-hit, entry-level approach that includes tools such as rules engines and screen-scraping. RPA is sometimes referred to as "macro-based" automation. Class 2, or "learning cognitive automation," encompasses capabilities beyond RPA, such as natural language processing. Chatbots fit here. Class 3, or "reasoning cognitive automation," represents true artificial intelligence, exemplified by IBM's Watson system.
Agencies are finding that class 1's RPA software bots are valuable for automating mundane tasks, such as cutting and pasting data from one system to another. These tools can work with existing IT systems, making them a great way to get started. The key is to start small: Decide what particular process to automate and then develop a proof of concept around it. From there, plan to scale the process across the organization and then move on to more complex tasks.
The use of chatbots (class 2) continues to grow because they can deliver not only cost savings but also a better constituent experience. One example involves call centers, which typically handle a high volume of fairly simple, repetitive communication tasks. Agents handling calls must toggle between various systems and manually enter information. Because of these distractions, the caller's experience can be less than ideal. Chatbots can free the agent from these repetitive tasks as constituents engage with the agency through cognitive-powered text or voice chat. Agents can devote more time to developing customer-centric skills. Los Angeles and Kansas City are among many governments finding ways to use chatbots to handle a variety of tasks.
Class 3's artificial-intelligence approach requires a substantial investment of time and resources but can address an agency's most complicated problems. Cognitive software mimics human activities such as perceiving, inferring and gathering evidence, hypothesizing and reasoning. The power comes in its ability to ingest huge amounts of data and quickly arrive at possible solutions -- something that would be too complex and time-consuming for humans. Juvenile-court judges in Montgomery County, Ohio, are testing Watson's ability to analyze reams of data in complicated cases.
For all of the promise of intelligent automation, agency leaders should realize that organization culture can be a barrier, so workforce and governance issues must be considered before plunging in. CIOs should plot strategy with the business side of their organizations because the introduction of IA is much more than a technology discussion; it can lead to a transformation in the way an agency performs its functions.
Leadership should carefully consider the impact on workers, particularly those who typically handle the rote work that a bot might replace. Communicating the benefits to these workers is key. They should understand that this change will allow them to focus on more value-added, strategic work, such as engaging with constituents. And while some tasks may no longer be handled by humans, new tasks such as bot configuration and management will take their place. The use of IA can lead to a surge in employee development.
IA offers government agencies huge potential to transform the way they provide services to constituents. But as with any transformation, leaders will need to obtain buy-in from across their organizations. Careful planning and clear communication are essential.
