I'm sure you've experienced spending hours doing the same thing over and over again at your computer, transferring data from a website to an Excel spreadsheet or sending identical emails. To solve this problem, Microsoft has launched Power Automate, a tool that allows us simplify routine processes without us having to be programming experts, allowing the software to work for us while we have a coffee.
This system is especially useful for those looking to connect different environments, such as performing tasks between Google's operating system and Windows 11. Basically, it involves create workflows that execute sequences of automatic steps, which translates into a huge improvement in productivity and a considerable saving of time in those tasks that are, frankly, a drag.
Installation and setup of browsers
For Power Automate to take control of your navigation, simply installing the app from the Microsoft Store on Windows 10 or 11 isn't enough. It's essential configure the appropriate extensions in the browser you use. Whether you prefer Google Chrome, Microsoft Edge, or Mozilla Firefox, the system needs a "bridge" to communicate with the web.
In the case of Chrome and Edge, the easiest way is to access the extensions menu and enable developer mode To drag the extension file located in the program's installation directory on your C drive, you'll need to do this. If you're using Firefox, you'll need to go to the add-ons manager and install the file manually. For everything to run smoothly, it's recommended to... disable background app running in the browser's system settings, thus preventing the flow from getting stuck.
Interaction methods: Extensions vs WebDriver
There are two main ways to make Power Automate interact with web pages. The first is browser extensionwhich is the default option and the easiest to configure. However, for more advanced users or specific environments, there is the WebDriver method, which offers a more technical control alternative.
If you choose WebDriver, keep in mind that it requires installing an executable file compatible with your exact browser version in a specific LocalAppData folder. Unlike extensions, WebDriver has some major limitationsFor example, it cannot be associated with windows that you have manually opened, but only with those that the flow itself initiates, and it can cause problems if you try to run several automations in parallel that require the window's focus.
Advanced web automation actions
Once the browser is linked, the range of possibilities is enormous. We can start with the basics, such as Start a new instance of Chrome or Edge, navigate to a specific URL or even create new tabs. The system allows us to interact with the page by emulating human activity, such as clicking on links or fill in text fields automatically.
For more complex cases, where the website's JavaScript doesn't react well, Power Automate allows you to activate the physical interaction of the cursorThis means that the mouse will actually move over the screen to press the button, although this requires that the browser window always be in the foreground and cannot be minimized during the process.
- Data extraction: It allows you to capture single values, entire tables, or lists from a website and save them in variables or directly in Excel spreadsheets.
- Form control: We can check boxes, select options from drop-down menus, or press radio buttons according to the flow logic.
- Conditional logic: By using the "If the web page contains" action, the flow can decide whether to execute one step or another depending on whether a specific text or element appears.
- Download management: There is a specific action to click on download links and assign the destination folder where we want the file.
Automation of complex processes and data
Imagine you need to audit the accessibility of several websites using a tool like WAVE. You could configure a workflow that reads a list of sites from a SharePoint or Excel fileVisit each page, wait a few seconds for it to fully load, and then run the analysis extension. After obtaining the errors, the system can transfer that information to an Excel spreadsheet and upload the final result to the cloud so that Power BI can then process it.
To achieve this, the following are used: CSS selectorsThese are like the exact addresses of the elements in the page's HTML code. If the website has several pages with the same structure, we can configure the paging system so that the robot continues navigating and extracting data until there are no more records to process, thus handling large volumes of information without human intervention.
Other connectivity tools
Not everything revolves around the browser. Power Automate also allows communicate with web services through HTTP actions. This is vital for those who need to interact with APIs, allowing them to send or receive data using GET and POST methods without needing to open a browser window, making the process much faster and more efficient.
Furthermore, the tool is capable of performing automatic screenshots of specific elements or the entire page, saving them as image files in various formats such as PNG or JPG. This is ideal for generating error reports or visual verification that a process has been completed successfully.
