Robotic Process Automation (RPA) is an alternative, easy way to automate repetitive routines. The word "robot" in the RPA term refers to a software robot that simulates the actions of a single user on interfaces. Modern integration platforms - such as FRENDS - have been doing process automation for 30 years directly under "interfaces", directly calling system interfaces.
However, RPA is just human simulation - it's thousands of orders of magnitude slower than a visually definable workflow that is eventually translated into code. RPA is also susceptible to failure, as changes to the browser-based cloud interface interfere with recording. Integrated process automation on the integration platform achieves the best results compared to the same automation with RPA alone. Sometimes, however, there are a lot of logic inside existing applications behind the user interface and moving it to the integration platform is a big expense. Instead of automating the process, we can automate simple routine tasks previously performed by humans by simulating a user - this is where the integration platform process automation performs RPA robot recording as part of the automation.
The RPA market is currently in a period of intense hype. The hype is shifting to real utilization of RPA and often over-exploiting RPA applications - for example, using RPA many times slower than interface automation, even if an interface exists.
There are no interfaces in the source or target system
The process to be automated requires more user interfaces, e.g. Excel Excel on screen
The business logic of the source or target system is coded at the push of a button on the user interface.
All of the automated steps occurs inside single User Interface (e.g. automating steps inside ERP via UI)
The source or target system vendor charges the interface separately for the destructive price of the interface process automation return of investment.
Intelligent Process Automation (IPA) is a combination of machine learning and process automation. Sometimes the process needs to change its operation based on past cases. A pre-taught machine learning engine in cloud platform provider may act as an decision engine for an automated process. An example of this is the automatic pricing of fuel at service stations based on competitors' price, time, location, total station coverage and other parameters. By entering these parameters into a machine learning engine (Azure Neural Network Regression), the automated process can...
update ever-changing teaching data for AI such as margins and prices
during the process, choose gas station-specific prices several times during the day based on the suggestion made by AI
HiQ is a proud partner of UiPath.
Interested? Contact us and we will give you a "RPA done Right" presentation.