BPM.com
  1. Peter Schooff
  2. Sherlock Holmes
  3. BPM Discussions
  4. Tuesday, 31 July 2018
  5.  Subscribe via email
What would you say is the best way to start a robotic process automation project to maximize success?
Accepted Answer Pending Moderation
White Rabbit: 'Where shall I begin, please your Majesty?'

'Begin at the beginning', the King said gravely, ' and go on till you come to the end: then stop.' [Lewis Carroll, 1865]
Comment
  1. more than a month ago
  2. BPM Discussions
  3. # 1
Accepted Answer Pending Moderation
If it's your first project (or an early project), take time to identify work that will actually be most suited to RPA. Not just technically suited, but operationally suited. Work that's routine and rules-based and relies on patchworks of systems, yes; but also places where a significant amount of very similar work is carried out in at least a small team. Otherwise it might be hard to get a decent return. Business functions like accounts payable are a great place to focus for precisely this reason.
Whether it's an early project or not, the same things that apply in a BPM context apply here: build a cross-disciplinary team with business involvement, work collaboratively and in the open, with users involved, work iteratively. Don't try to get to 100% as in almost all cases this will be a fool's errand.
Comment

Good advice, totally consistent with the method of choice for Strategic Planning in any corporation where the challenge is more "promising" initiatives than can be undertaken in any time slice, because of scarce resources.

See . . .
"Strategy Development – Same old, or entirely new?" 2018-06-10

https://wp.me/pzzpB-Sd
  1. more than a month ago
  2. BPM Discussions
  3. # 2
Accepted Answer Pending Moderation
Find the appropriate use case - tasks that are highly manual, redundant, structured, and don't vary.

For example, in order to justify annual awards, government staffers manually collect the same data from the same various systems to enter into a same payroll system. There are 10 staffers handling roughly 13,000 actions with each action taking 5 minutes. So roughly $40K of manual work in this one task area that can be eliminated by using RPA bots to gather, consolidate, compare, and post data.

Now you need to gather requirements to create a POC.

Questions to ask:
-What are the applications used (Java, SAP, Web, Excel)?
-What are the screens in the applications that contain the data?
-What are the data to collect? Structured? Unstructured?
-Does the screen configuration change?
-What are the exceptions? How handled?
-What are the outputs?

You get the idea.

RPA is straight forward. The bot replaces the person. Program the bot on where to go, what to see, what to get, what to do with it, and where to put it.
Comment
  1. more than a month ago
  2. BPM Discussions
  3. # 3
Accepted Answer Pending Moderation
As often happens, the initial excitement is often to get a bunch of "IT" work done with cheap, relatively unskilled labor who doesn't understand your business. OR, to have your business folks do the work (cheap as in hidden cost, not a hard cost on the P&L). Those approaches typically don't work well. It isn't that these personas can't be enabled, but that the fact that an organizations goals are to foist this work on these people is a sign of how little commitment to investing in the organization there is.

So take a different approach. Get some experts in. Address automation in context of the business, understand the pitfalls and what makes automation brittle. Realize that most people are just there to take the order, do the minimum, and exit. Go get the real business value that is there to get (in a couple of our projects the returns have been eye opening good). I heard one company describe RPA in their organization as "instant technical debt" - which is a good indicator of how it can go wrong, or how it can be used as a band-aid to get quick resolution on a bit of integration or automation to be replaced by something robust later (which may still be RPA, but more "designed" in nature). But it also can be used to unlock value in areas that are otherwise stubborn to change.
Comment
  1. Ian Gotts
  2. 4 months ago
  3. #5570
So true.
*mic drop!
  1. John Morris
  2. 4 months ago
  3. #5572
+1 mic drop!
  1. more than a month ago
  2. BPM Discussions
  3. # 4
Accepted Answer Pending Moderation
Simple. 1. Understand the business processes. 2. Gather requirements. 3. Build bots. 4. Document what you've built so you can improve the business with what you've learnt. So back to 1. and round the cycle.
Comment
  1. John Morris
  2. 4 months ago
  3. #5573
You have a knack for the MVE ("minimum viable explanation"). :)
  1. more than a month ago
  2. BPM Discussions
  3. # 5
Accepted Answer Pending Moderation
RPA Projects should be kick-started based on the following 3 Steps:

  • Relevance of the Process/Project from a Customer viewpoint
  • Technology Implementation feasibility & customization effort if any from an IT standpoint
  • Benefit & RoI Calculation from a Business standpoint


The outcome of the above exercise can lead to "Identification of the right RPA candidate". There can be multiple ways to arrive at this via heat-map analysis or exhaustive data-driven metrics.

It's important to have the right strategy aligned and channelized for the FIRST RPA Project - as it may set the foundation stone for many others to follow.

Typically, RPA projects are considered as tactical solutions which stitch together applications built across discrete tech stacks. Hence there will be always a maturity roadmap closely associated with RPA Projects

  • The process the RPA flow is wrapping can go through an optimization/re-engineering exercise
  • The legacy system/green-screens used can get enlightened with a digital transformation
  • The multiple hand-off processes can be streamlined with seamless Integration enabling STPs (Straight Through Processing)


So, the starting of an RPA project can never be considered as an end of exercise/automation - it just gets better with each iteration!
Comment
  1. more than a month ago
  2. BPM Discussions
  3. # 6
Accepted Answer Pending Moderation
It's no difference from any kind of transformation project - sign the contract with a highly motivated customer.
They will help you then by:
- identifying the right task candidates for RPA
- sustaining the change management effort
- driving organizational adoption
- transitioning / transforming redundant workforce

If your customer is only motivated by the IT fad around the transformation project, it will fail miserably.
CEO, Co-founder, Profluo
Comment
  1. more than a month ago
  2. BPM Discussions
  3. # 7
Accepted Answer Pending Moderation
Working mainly on business process optimizations and automations, for us at NSI, it has been natural to start RPA initiatives as a part of larger BPM projects.
In each BPM implementation we usually get to the point of defining with which applications a front facing, operational business process has to “talk” to in the back, in order to harness the largest possible impact in terms of improving response times and data quality. It is that point, where the decision between using a web service vs. training a bot, kicks off the accompanying RPA undertaking.
A specific example would be for a banking process, automating an export financing process as part of the corporate banking operations. One eventually would get to the point of having to interact with the international money transfer system SWIFT and decide whether to use a web service or a bot in order to interact with the overlaying business process - and a bot would be normally a better choice here.
NSI Soluciones - ABPMP PTY
Comment
  1. more than a month ago
  2. BPM Discussions
  3. # 8
Accepted Answer Pending Moderation
I think there is a need to think Business Process Automation where RPA is just one component. There needs to be caution where RPA needs to rely on existing UIs and legacy change and do not not rely on vendor promotion! Think of bigger picture which BPM can quickly facilitate and do research on capabilities that can deliver quickly direct from business input. Always test before starting a project....the no low code movement is growing and should help with understanding just how....
Comment
  1. more than a month ago
  2. BPM Discussions
  3. # 9
Accepted Answer Pending Moderation
Because such projects are digital transformations then do not forget to invite an architect, please.

And remember what may happen without architects - "Lidl software disaster another example of Germany’s digital failure" [1].

Thanks,
AS
References
  1. https://global.handelsblatt.com/companies/lidl-software-flop-germany-digital-failure-950223
Comment
  1. John Morris
  2. 4 months ago
  3. #5574
They had a lidl problem . . .
  1. Kay Winkler
  2. 4 months ago
  3. #5581
:d
  1. more than a month ago
  2. BPM Discussions
  3. # 10
Accepted Answer Pending Moderation
Although popular and widely discussed recently, RPA by definition is very limited technology, not a panacea on IT problems existing in a company. In this capacity, RPA is senseless and dangerous when applied alone.

Technically, RPA is a fragile collection of scripts, which are difficult to impossible to adapt on even slightest change in business environment. An intent to rely widely on RPA is equal to an attempt to conserve indefinitely a status-quo with legacy IT systems. Evidently, it is only a matter of time when such a crippled construct will collapse and bury business.

However, it does not diminish an important role of RPA as a facilitator and accelerator of digital transformation through robust instant automation of routine legacy procedures. It should serve as a proof of concept in discovery of business transformation potential in an organization.

To reap all benefits of this automated discovery, it is essential to explore RPA only in the context of a business model where RPA will allow to quickly evaluate transformation potential for previously identified bottlenecks and inefficient processes. Therefore, always start with careful business mapping and solid BPM analysis to succeed with RPA.
Comment
  1. more than a month ago
  2. BPM Discussions
  3. # 11
  • Page :
  • 1


There are no replies made for this post yet.
However, you are not allowed to reply to this post.