On 14th September 2019, Gleematic founder – Christopher Lim was honored to be invited to give an insight speech at Fudan University, Shanghai – China. Christopher Lim gave a knowledge-sharing speech on RPA/AI design and development.
Robotics Process Automation (RPA) arises as a solution to operational challenges in the business. Its roadmap is driven by the interest and pain points of target markets in addressing their operational challenges. The unfailing challenges finally led to the design and development of RPA.
The concept of RPA is automation without programming, although RPA was originally based on screen-scraping. However, because screen-scraping was only for programmers and there was a criterion for this technology to be widely used and deployable for a wide range of users, RPA then arose as automation without programming.
Enhanced with Artificial Intelligence, RPA to be called RPA/AI or cognitive automation, can do more than just automating work. The presence of artificial intelligence allows software robots to do human decision-making tasks that require knowledge, such as predictive or sentiment analysis.
In cognitive automation adoption, the design and development method used is important because it can affect the whole process and the product outcome. There are two kinds of development we know: waterfall and agile. But what is the difference between the two? How are they different from each other? Which one brings more advantages and disadvantages? What kind of project it’s suitable for?
Waterfall development methodology works in sequential order, the project development team moves to the next phase of development only if the previous step is completed successfully. The testing of the product starts once development is over. Meanwhile, agile development works differently.
The agile development methodology is a continuous iteration of the development and testing process. In an agile method, development and testing are done concurrently. Basically, in this agile method, the project development team completes the whole product first, testing or re-development of steps are done concurrently. This method allows faster and effective development, also provides two-way communication between customers and developers.
However, there are advantages and disadvantages between waterfall and agile development. The waterfall method is the easiest model to manage during project development and has a faster delivery of the project with a well-documented process and results. However, it is not an ideal model for a large size project. If the requirement of the product is not clear at the beginning, it will affect the whole project and therefore it’s a less effective method to implement. Since the testing process starts once development is over, it’s rather difficult to move back to make changes or fix the previous phases.
Agile development brings more advantages to the development of a product. The agile method is a bridging process between client and developer, as it involves the client continuously in every stage of development. The agile development method is based on incremental progress so the client and developer team know which stage or what is and isn’t completed, so the quality of development is maintained and the risks that could occur in the process can be reduced. However, this method is not useful for a small development project. The project can easily deviate if the project manager is not clear of what the outcome is desired to be.
For end-users, it’s also important to consider few things in implementing cognitive automation for your business operations: type of process or workflows, the latency of applications, budget, resources, time, and the number of transactions you want to process automatically per unit of time.
In a nutshell, cognitive automation is indeed a solution to help many business practices overcome and address their challenges, but also it takes more attention and knowledge in designing and developing the automation software itself so that it can work effectively and meet your desired outcome.