Imagine that you are working as an accountant—for example, in a shared service center—and your personal robot is doing your daily (repetitive) job for you.
Might this become reality? In fact, it is already possible.
What robots can do today
Today, robots can be installed on top of an ERP system, for instance, to conduct repetitive tasks 24×7. These robots, which can learn such tasks as easily as humans can, generally work like recorders based on a graphical interface. Tasks such as normal posting or creating an asset need to be conducted only once for the robot to learn. After that, the robot will repeat the task as often as necessary based on the learned steps and the underlying IT system.
The advantages are obvious:
- Short implementation times with manageable effort
- Almost no human resources required for repetitive tasks
- 24×7 task performance
- High flexibility thanks to fast, easy setup
What robots can’t do today
What is the downside of this technology? First, robots typically work on the user interface level – the surface of an IT system. That means that if the underlying system changes, the robot must also adapt or be replaced.
That said, robots use only the existing system with all its functionalities and restrictions. The robot can work only as well as the underlying system allows, so if the system setup is poor or includes many restrictions it will try to mitigate these limitations. The same consequences arise with unharmonized and standardized processes, where separate robots might be necessary.
From an IT architectural perspective, there are many questions that must be answered:
- IT design – What is the best way to embed robots into an overall architecture?
- Security – How should security be handled?
- Control – Should everybody be able to install their own robots?
- Maintenance – How can robots be included in an upgrade and maintenance strategy?
Although robots bring both advantages and unanswered questions, the trend in finance seems clear: to maximize automation of process operations.
But robots are not the only way to achieve automation. Why optimize robots only on the surface level of processes and IT applications? Isn’t it more promising and sustainable to automate directly in the core of applications via predefined rules, and in the future, via machine learning and artificial intelligence?
A catalyst for further automation
While this approach is much more complex and time-intensive, robots can serve as a first step and a catalyst to drive real automation in finance.