Jesse Lambert & Lee Plumb
You know AI has officially gone mainstream when it’s a hot topic within the traditionally risk-averse space of Federal government acquisition. Yet that was the case at the ACT-IAC 2019 Acquisition Excellence conference in Washington, D.C. in March, with panel discussions titled “Intelligent Automation and Artificial Intelligence in Acquisition,” and casual references to “RPA” (Robotic Process Automation). And it’s true that several government agencies not commonly associated with cutting-edge technology (e.g., GSA, HHS, IRS) are automating business processes to harness efficiencies and cost-savings. One speaker at the conference noted that automation could save the Federal government $40B over just the next three years.
Still, the truth is that the government is still in the early stages of AI application, currently focused more on automating transactional process than advanced machine learning. The process design approach is critical to ensure that an automated process is most effective and with the support of the people affected by it, going beyond just the people manually performing the process today. As we progress from these early automation initiatives to more sweeping adoption of AI, here are a few some potential hurdles, and how Evans Incorporated uses Human-Centered Design to inject real intelligence into process automation:
- Near-term: Agencies and most consultants already seem to embrace that the people currently performing the work to be automated need to be involved during process and “bot” development, but that doesn’t equate to “human-centered design”. It isn’t enough to just work with staff to map out the process and design something that replicates it or improves it slightly. This is the time to fundamentally improve the function supported by the process, and to think ahead by considering qualitative requirements, such as ethics, compliance and treatment of sensitive Personally Identifiable Information (PII) data.
- Evans’ capabilities in design thinking, business process improvement, and change management pair well with AI to capture the nuances of process being automated, improve upon it where possible, create a foundation for evolution, tailor outputs to the needs of decision-makers, and adapt the organization to enhance adoption.
- Medium-term: The current automation target at most agencies is administrative and transactional work, freeing staff to focus on higher value responsibilities, but the attention is mostly internally-focused. As government expands automation to externally-facing processes that impact the public directly, agencies will need to consider a more diverse stakeholder environment, as well as policy, governance, and privacy issues that will surface when technology disrupts culture.
- Evans emphasizes stakeholder engagement and communications as core to any organizational improvement. We can help clients implementing AI solutions define their stakeholder base to better understand the users of their automated systems and the impact of those systems on other downstream stakeholders, then incorporate their needs into service design.
- Far-term: As government experience the efficiency gains of automation, many employees will transition to supervising automated systems and performing functions that are less conducive to automation, but positive outcomes from this transition aren’t guaranteed. Can employees whose primary role has been replaced by “bots” be retrained for higher value work, or will they become unemployed? Agencies will need to redefine the workforce and create more high-value business functions.
- Evans’ human-centered approaches to strategy and organizational design enables clients to build their future workforce, harnessing the power of AI to elevate their capabilities. We interpret organizational goals through the lens of human capital, then craft a talent management plan to achieve those goals that aligns with and reinforces positive cultural attributes.