Robotic Process Automation

Robotic Process Automation, or, RPA, a trend I hardly followed till now. Looks like there is much to explore and see how AI and ML are being put to use for everyday jobs.

Some tools and resources to explore:

Getting Started

  • Adapt automation to your mindset. Start a habit to look for tasks that are repetitive, high-volume, rule-based and prone to human error.
  • Apply this 4-step approach:
    1. Start to automate micro-tasks first. The simpler, the better.
    2. Look for processes where you can gain a lot of value by automating.
    3. Map out the processes step-by-step. Then create the automations and test them.
    4. Measure results. See if the initial ROI justifies larger investments.
  • Decide whether you want to go the commercial or the open-source route (see Tools-section below).
  • Remember also scripting, macros and APIs. RPA is just one of the tools in the automation stack (see Tools-section below).
  • Employ professional service providers to create automations for you. Just Google RPA service + country and you’ll probably find dozens. No need to do everything yourself.

Pros

  • RPA lets you start your automation journey with a low-risk/high-reward approach. Integrating existing software systems might be expensive or even impossible. RPA lets you automate legacy systems with a bootstrap budget.
  • Community editions of software let you get started for free.
  • Software robots will become commoditized in the future. With increasing competition, prices will be driven lower. As prices go down, RPA will extend its reach from big enterprises towards small-to-midsize companies.

Cons

  • Immediate resistance from your team is likely to occur towards automation. Onboard the team and management early on to embrace the change as it frees their time for more valuable tasks.
  • Typical other pitfalls: automating processes that are too complex, difficulties to scale and selecting processes that have insignificant business value. More lessons learnt can be found from research.
  • When scaling further, investments are required. But there’s always the open-source option.

Cases

  • Wärtsilä got their first automation running in 3 weeks. They have now 400+ processes automated and over 5,000,000 € saved.
  • Coca-Cola automated 50+ processes across multiple SAP systems. The company found 16 extra hours/day that could be used for productive work by the new digital workers.
  • Posti (the Finnish postal service) used RPA together with machine learning to automatically process 3000 purchase invoices monthly.
  • First Home Bank used bots to help process over 6,000 PPP loan applications in a few months. The bots were 30x faster than humans.
  • Lyse automated submitting applications for government approvals. 20,000+ work hours saved annually.

Resources

Takeaways

  • Pilot automating with a low-risk, low-cost approach. Build micro-automations first and measure success. Continue from there, IF feasible.
  • Let professionals develop enterprise-wide, mission critical automations. Utilize trained citizen developers for creating simple automations at scale.
  • Ensure that your organization embraces the change rather than fighting back.