Handling a robot – how to take care in order not to be careless
- Building a robot is usually not very difficult.
- A real challenge is to control the engineering life cycle of a robot (from idea through implementation to maintenance).
- A monitoring of robots’ work is necessary even in a very little scale of operations.
- The most important aspect is acceptance of the robots by their human co-workers.
How easy is building a robot?
The modern RPA platforms promise that creating robots might be as easy as creating a table in Excel or a presentation in PowerPoint. Thanks to the use of templates, ready-made components or even recording tools, you can actually create a working robot in a short time. Besides, this kind of robot might perform a useful job. It seems that we should only search for another processes and automate them successfully.
Creating a robot is not enough
Unfortunately, as always, the reality differs from such a perfect example. Creating a robot in a non-trivial business case requires an experienced RPA developer, but this aspect is not the biggest challenge. The hardest part is what comes next, which is the ongoing work with the robot, ensuring that it does its job flawlessly every day, and if there are any changes, its code will be adapted quickly and efficiently.
Robot is a software
First of all, you need to realize that a robot is a piece of software that is subject to the same engineering rules as the code of a website or windows application. In a short blog post, there is no place to describe the entire process in detail, so I will focus on the most important ones, i.e. change management and monitoring of robots’ performance.
Robot’s life cycle
We must be sure that the change management process is in hands of a person or a team that has knowledge and competences in order to take care of it in its entirety. If even a small part of the process is not covered or the responsibility for it is not clear, sooner or later the robots will simply stop working.
Usually, such a process is in the hands of a dedicated competence centre, or is part of a larger IT-managed process. Such team makes sure that only authorized persons can submit changes, so that such changes are put into execution in an orderly manner, and that they appear in the production environment on a specific schedule.
Apart from the record of changes, it is very important to manage the version of the robot’s code. You cannot allow a situation where different copies of the robot are circulating in the organization and it is not clear which one is currently working. This is especially important in an emergency where every minute counts. Of course, we also have to make sure that selected people can modify the code, and not to lead to a situation where a single person would have the right to make the modification without a review by another person. It is absolutely critical that every change is subject to rigorous testing. To do this, you need to effectively manage separate development, test and eventually production environments with separate permissions. The optimal solution is to use both automatic testing mechanisms and automatic code migration between environments.
These are all very technical issues, but mastering them is necessary for robots to run reliably and to be safely entrusted with even the most critical processes in the company.
Robots’ work monitoring
Sooner or later, even the best-tested robots will encounter critical failures. In such situation, it is necessary to ensure that the information about the stoppage reaches the people concerned as soon as possible. While the development of small-scale robots can be done without advanced tools or processes, it is difficult to imagine that the monitoring is performed “manually” or “by watching it”. If the robot’s creator is supposed to supervise the work by looking at the monitor screen, he probably wouldn’t have had time for any other activity.
Without an automatic monitoring and reporting system, even the smallest installation cannot be imagined. In order for the entire system to work properly, it is first of all necessary to ensure in the process of producing robots that they will be able to clearly inform about their condition and send alarm signals precisely and only when human intervention is necessary. Robots presence makes sense only if they decrease human labour input. If robot informs about its condition too frequently, it will soon be ignored, and consequently turned off. On the other hand, designing an intelligent notification and signal interpretation system will allow the maintenance staff to react quickly and get the robots back up and running.
The best RPA platforms usually offer embedded modules for monitoring, or external tools that are used especially for such tasks. Certainly, embedded tools will be better suited to the specificity of robots, while external tools offer wider functionality. It does not matter which of these variants we choose, as long as we plan well what we want to research and how to react to exceptional situations.
Work of hybrid teams
Robotization still arouses a lot of emotions, especially among employees whose current tasks are to be delegated to robots. If we do not care in the very beginning of the project about the proper communication and an idea for a future work of people and robots, even the best technical solution will not be effective. There are some rules that must be followed in order for hybrid teams to be successful. First of all, find the tasks that frustrate your employees and that are the least likely to be done. These are the best candidates for robotization. On the other hand find tasks that are interesting, creative, and empowering employees, but they don’t have time for them right now due to the workload of ongoing tasks. My experience shows that in each workplace you might find both categories without any problem. This means that we will have a good plan to reallocate tasks when robots arrive. Then such robots should be delivered as soon as possible and employees should be shown how their everyday life is changing for the better. If we take care of these aspects, we will not only obtain more efficient and effective teams, but we will also free up time in which employees will, on the one hand, look for new ideas for improvement, and on the other, they will become true ambassadors of robotization.
It might seem that the most difficult challenge is building robots that work efficiently and perform useful work. Meanwhile, the most important thing is to convince people that robots can be their Digital Teammates, and then to control the entire life cycle of the robot, from the idea, through implementation, to the ongoing supervision over the work of robots.
These are not easy tasks, but the eventual benefits from robots’ work are so tangible that it is worth overcoming these obstacles.