Task Mining use on real life case

Help Task Mining
,
1

Hello everyone,

I’ve been exploring the Assisted and Unassisted Task Mining tools and I wanted to share my experience with the Assisted Task Mining Project. I ran some tests with two users and unfortunately, I came to the conclusion that this product is not very useful (or perhaps I don’t know how to use it properly). The flowchart was difficult to read and edit and the merge functionality didn’t work as expected - it merged tasks in a really awkward way.

So, I would like to ask the community if anyone has had success using Task Mining in a real-life case. Any advice or tips would be greatly appreciated.

Thank you!

2

Hi,
TM could be used in real-life scenarios and some tips for success:

  1. Process Discovery: Task Mining is excellent for discovering processes that are not well documented or understood. By capturing the actual work as it happens, you can gain insights into the real-life workflows within your organization.
  2. Complex Workflows: If the workflows are complex, Task Mining can help you identify patterns and commonalities that might not be apparent through interviews or observations alone.
  3. Improving Accuracy of RPA: By using Task Mining to capture the nuances of how tasks are performed, you can design more effective and accurate RPA (Robotic Process Automation) bots.
  4. Employee Onboarding: Capturing tasks can also be used for training purposes, helping new employees understand the workflows by providing real-life examples.

Tips for success with Task Mining:

  • Clear Objectives: Define what you want to achieve with Task Mining. Are you looking to automate tasks, improve processes, or gather data for analysis?
  • User Selection: Choose participants who are representative of the various roles and processes you wish to analyze. Diverse input can provide a more comprehensive view.
  • Effective Communication: Ensure that all participants understand the purpose of the Task Mining project and how to use the tool effectively.
  • Data Quality: The output is only as good as the input. Make sure users are performing their tasks as they normally would to capture accurate data.
  • Analysis: Spend time analyzing the data. Task Mining tools can generate a lot of information, and it may take some effort to sift through and find the valuable insights.
  • Merge Functionality: If the merge function isn’t working as expected, try to merge tasks manually. This will give you more control over the outcome and may provide better results.
  • Technical Support: If you encounter technical issues like difficulty reading or editing flowcharts, reach out to the UiPath support team or community forums for help.
  • Iterate: Use Task Mining as an iterative process. The first pass may not give perfect results, but you can refine your approach and run additional mining sessions as needed.
3

@srinivasmarneni

Quick question: as a professional RPA dev, do you use this tool in your work? And you see value in it?

I am asking because I have just finished testing the unassisted task mining (UTM), for which I have almost 70 hours of recording. I ran the analysis (yes, I spent 500 AI units for this), and the result was useless. It just found 6 tasks with good automation potential… but the six tasks are the same (login into a web portal), not unique tasks. To complete my disappointment with the result, it did not flag a high-value repetitive task that I know for sure the user was doing at least 1 hour a day during the recording period of 1 week.

So my conclusion after testing both tools, ATM and UTM, is that they don’t deliver what UiPath marketing promises. In the best scenario, ATM will help a business analyst start to draft the process map, and that’s it. And regarding UTM, I am not sure if it worth the cost to run the analysis (based on the test that I did)

4

Hi @Fred_Torres

You can try elaborate a strategy with you UiPath account manager to implement Task Mining successfully and get all the potential that it offers.

There is some aspects that may impact your analysis like configuration, recording users and recording time, the user knowing that you are recording etc.

At first time we tried the Task Mining we didn’t get a lot of potential processes for automation so now we are elaborating a strategy with our UiPath account manager for that.

5

Hi @Fred_Torres,

I have a contradictory view on this. I recently delivered 2 projects with task mining and they were successful. Important Pointers

  1. If your final task list is not having anything that means you have very less automation potential in your use case and you can analyse unprocessed data to analyze more using tableau.
  2. You should know how to calculate cost matrix to present it to client that this automation help with number of hours and cost they are going to save
6

@Fred_Torres There are few best practices and governance you need to follow if you want to see good outputs analysis reports for UTM. Please refer all the sections of this documentation and especially this link :https://docs.uipath.com/task-mining/automation-suite/2023.10/user-guide/best-practices-utm. It is recommended to use ATM as a default Task mining mechanism if you have access to SME and get an quick over view and UTM as an alternative if you want to capture a process that spans multiple hours and process steps …

7

Hi @Fred_Torres,

I successfully completed almost 15-20 task mining projects successfully. If you need any assistance with ATM or UTM let me know.

8

Hey, can you tell me any use case for assisted task mining, I need to create Assisted task mining assignment for users.