Standardization of the robot development process

In one of the previous blog post (https://dtmates.com/robot-manufacturing-process-automation/) I wrote few words related to the automation of the robot manufacturing process. One of them is the concept of standardization. Here I would like to share few words on – why standardization is needed and why we do it at all.

What we try to standardize and how we do it:

Robot standardization

All of our robots share the same pattern; they are made of modules responsible for different aspects around business processes (build the same way for all robots) and main module responsible for realization of the client’s business process. We also maintain the naming standards for elements of both the robot and the platform itself. 

Standardization of the process’ logical division

All business processes are divided in the way that lets us maximize the speed of robots’ work and allows us to scale robots horizontally. Furthermore, this divided business process makes change management easier and less prone to errors.

Configuration management standardization

All of our robots use our central configuration management system and the configuration itself is used and stored in a standard manner.

Testing process standards

Robot’s testing process preparation is an inseparable element of all robot shepherds’ work. For all of our robots we create automation tests performed by other dedicated robots.

Reporting and monitoring standardization

Our robots always generate reports in which they describe their work – these reports have business character and contains data related to the business process. In addition, all operations are logged and stored in external log repository. It allows us to monitor our robots on-line, both in terms of business and technical.

Why we do it? Why we standardize at all?

The quickest answer is – because it is faster and cheaper (though at the beginning it seems the opposite). Thanks to standardization, new robo shepherds are able to start their work faster and can deliver robots for our clients earlier. Our standards facilitate the change management process, shorten the time needed to find and remove errors and let us automate robots’ installation. Finally, it gives us more control over individual robots, which in turn allows us to build entire robotic farms.

Standardization is a process, it doesn’t mean that once created standard is eternal and carved in stone. In our case, we make reviews of our standards from time to time (we discuss their form) and if we decide that some element should be changed, we prepare a new version. However, such a change must be well-argued and can not result from the short term needs (“taking a shortcut”).

We know for experience that taking shortcuts always ends with unexpected problems.

Bartosz Pietras
Head of IT

When do I start to pay for robotization?

If you decide to go for robotization on your own, most often with a help of external specialist who assist in rolling out RPA project, than significant up-front investment need to be done. Apart from the actual cost of hiring external company (and these costs may exceed hundreds of thousands Euro), internal robotization is associated with following costs:

  • Investment in IT infrastructure required for hosting robots internally
  • RPA licenses – in most vendors’ offer licenses are acquired in a long term agreements (exceeding 1 year). Moreover, one license can accommodate robots doing equivalent of 3-4 people work, what means that efficient utilization of RPA licenses requires scale.
  • Investment in IT Center of Excellence team which will be able to maintain the infrastructure, platform and tools required for robotization.
  • Setting up RPA development team, which will not only develop new robots but also reconfigure them when needed after they are launched.

Digital Teammates® is offering robots in pure Rent-a-Robot mode so expects to be paid for robots only since they start doing actual work for our Clients. The whole phase of potential recognition, development, testing and deployment is investment- and cost-free for our Clients. So all the above costs associated with launching robotization internally are non-existing in Rent-a-Robot model.

Konrad Jakubiec
Board Member

How much does the robotization cost?

Largest organizations can opt for fully fledged robotization setup. That means that recruitment of full RPA Center of Excellence is needed. On the top of that a team of RPA developers, IT infrastructure and licenses need to be acquired. This approach is recommended for organizations which have a potential for robotizing hundreds of FTEs jobs. One of very important factors of estimating costs of robotization in this setup is where the Center of Excellence is located. That is particularly important for international organizations which tend to have R&D and IT competences located in more developed countries and back-office operations outsourced to places with lower employment costs. Such setup increases pressure on robotization and requires even larger scale as development of robots is more expensive relative to back-office costs and achieving economic benefits is even more challenging.

Typical smaller robotization projects last between 6 and 12 months and consist of strategy, process analysis, process optimization and robotization potential identification. From our experience this approach is the trickiest as the further robotization and maintenance of robots requires having RPA developers and part of Center of Excellence on board. That means that without careful prior planning robotization in such mode may miss initial economic analysis.

There is another option available on the market which is based on Rent-a-Robot business model. Robotization with Digital Teammates® is done this way, what means that it doesn’t cost anything before robots start doing actual work. Their ‘salary’ is a fraction of the cost of an employee, who did the same work before robotization. Typically, robots start their work 4-6 weeks since the start of cooperation. What is even more important, in Rent-a-Robot mode Clients do not need to invest in RPA competences in order to have fully functional team of robots on board. In such mode robotization is considered as a straightforward saving from day one rather than a cost.

Konrad Jakubiec
Board Member

Robot as a threat or opportunity for job market?

In our view – definitely an opportunity.

We have a very consistent observation: in our experience way to many people worldwide do the tasks which are not made for humans and fit perfectly well for robots. We mean tedious, repetitive and error prone activities that very often yield decrease in motivation in teams. Every single management we have talked to can think about hundreds of better and more creative things people could be doing should they not be forced to spend so much time on tasks that robots like so much.

We strongly believe the more robots are deployed in workplaces the more satisfied and motivated people will be at work. And in foreseeable future we do not expect a situation where robots might start posing a threat to humans on job market as for many years we will live in a paradigm where there is way to much work to be done.

Digital Teammates thanks to having robots on board could develop a team of Robo Shepherds – team responsible for developing and maintaining the robots. As robots are deployed in operations we could provide an opportunity to people who did not have any IT background to start working with technology. Such opportunity would never be possible without robots. This is the path we expect the robotization will open – people will have more opportunities to realize their creative potential.

Konrad Jakubiec
Board Member

What are main obstacles to robotization and how to overcome them?

In one of my previous articles I discussed main reasons to go for robotization – https://www.linkedin.com/pulse/what-main-reasons-why-companies-go-robotization-mariusz-pultyn/ . Now I want tackle a more challenging topic – an analysis of main obstacles companies have to overcome to succeed with robotization.

  1. Why should I bother ?

Although robotization can be now regarded as a mainstream tool for enterprise digital transformation it is sometimes difficult to find a reliable source of information on tangible benefits robotization can bring. There are obvious reasons like reduction of the cost of operations, improvement of quality and efficiency and quite many non-obvious that I discussed in more detail in previous post: https://www.linkedin.com/pulse/what-main-reasons-why-companies-go-robotization-mariusz-pultyn/

In short, main reasons that companies should be driven by in making a decision about robotization are:

  1. It is difficult for a company to automate to the required extent with IT, and not all automation is economically justified that way
  2. Company is looking for quick wins against labor market challenges (attrition, rising wages, training, etc.)
  3. A will to release employees from the most tedious and repetitive tasks
  4. Increase in efficiency
  5. Improvement of quality
  6. Taking a first step to try and experiment with AI

I’m sure that reading that post will answer the question of why should you bother with robotization. If you are not fully convinced – please let me know in comments of address me directly.

  1. Lack of knowledge of assessment of robotic potential

The general rule of thumb says 20-30% of all office work is in range of robotic technology, but actual potential can be much more than that depending on your specific application environment. There is very short questionnaire that can help you identify processes that are best for robotization:

  • Process is highly standardized, documented and stable
  • Input data are digitized
  • Process requires significant amount of repetitive manual labor
  • Automated system is relatively stable in terms of adding new functionalities and in near future there are no significant architectural changes planned
  • Automated process is human error prone (requires large amount of data input, especially sensitive like amounts, dates or clients’ IDs)
  • Automated process covers few IT systems what requires an employee to be able to multitask and frequently switch between the applications
  • Automated process has significant volume fluctuations what makes adequate staffing challenging for Customer

If your processes share at least some of those characteristics they are most probably good for robots.

After such brief assessment you can go two ways mostly depending on your organization culture – either find some advisory help to analyze in detail all your potential upfront or start small with first process and robotize it and then build your experience in agile and iterative process of continuous improvement.

  1. Lack of robotic skills

Here you also have couple of choices. The most obvious one is that you may decide you don’t need robotic skills at all and let somebody else do the robotization for you, ideally providing robots-as-a-service. This is the best choice when robots are not your core business or your scale of operations does not exceed hundreds of employees. On the opposite you can try to establish Center of Robotic Excellence yourself by acquiring talents on the market or by asking for some consulting on how to establish and grow such competencies.

  1. Fear of implementing “temporary” solution that stays forever

Because robots work without changing of underlying IT systems and somehow cover inefficiencies of current user interfaces can be perceived as “temporary” solution until “real automation” is being introduced. The truth is that robots are fully fledged automation solution that in most cases has no alternative. This is because usually there are valid reasons that automated low-level interfaces haven’t been built or systems haven’t been changed due to sometimes prohibitive cost. On the other hand robots by their nature are extremely lightweight and can be a great bridging solution until those “ultimate” solution are being delivered.

  1. Potential collision with other digital transformation projects

As said above robots are now a part of the mainstream digital transformation tools and should be used as complimentary solution to other automation efforts. The best way to avoid potential collision is to simply put robots on your transformation roadmap after properly assessing their advantages over other activities. Big transformation projects usually take some time to implement, so robots are ideal bridging solution as I mentioned in previous paragraph. Robots are also perfect to fill some gaps in ergonomics of newly implemented solutions, especially when some difficult decisions on scope had to be made right before deployment of big new transformational project. Robots can give you some time to think about target employee experience when interacting with newly built systems.

  1. Substantial upfront investment

Even if robotic infrastructure is usually much less costly than typical IT solutions it still requires some servers, either on-premise or in cloud, some robotic licenses and above all skilled team to keep thing running smoothly.

There are some ways to overcome this issue – one has been briefly discussed above – you can simply find a provider of robotic services and pay only for the work robots will do for you. Then your upfront investment is close to zero.

You can also invest a little more and go for trial or free versions of robotic platforms with lightweight infrastructure in the cloud good enough to conduct Proof of Concept of robotic process automation and then try to scale it up. This approach is quite popular, but I wouldn’t recommend it, as results of small scale implementations can be very misleading when complexity increases.

Again there is a third way to tackle the problem – especially when your organization has a culture of creating more formal business cases for potential projects. In most cases properly assessed costs are greatly exceeded by the benefits of robotization, so upfront investment pays off in relatively short period of time.

  1. Difficulties of change management

One of the most important issues to be addressed before robots are introduced is what to do with employees’ time released by robots. This problem is less pressing nowadays as there is usually much more work to be done in mostly growing companies than employees able to process additional volumes. If that is a case you simply have to reallocate work giving people more interesting, creative and value adding work. When robotization is done in agile way the change is gradual and continuous, so people have enough time to adapt and usually come back with more ideas of what more robots can do.

If your robotization project is going to be implemented in more big-bang way or you are mostly motivated by reduction of your staff you might need to plan employee relocation as one of components of digital transformation of your organization. There is a plenty of advisory help you can find to help you deal with change management, however I would strongly recommend to go for gradual and iterative approach with robotic implementations, because it better aligns with robots’ lightweight nature.

  1. Fear of defeat

All new activities bear some risk of failure, but it is just stating the obvious. Why I’m convinced this risk is much lower with properly introduced robotic process automation ?

The most important reason is something I mentioned before – robots by their nature are aligned with agile paradigm. This means they can be implemented in much shorter times and with much less cost and much less stakeholders involved than typical automation activities. This helps to achieve positive feedback loop which provides that small robotization success leads to another small robotization success and eventually becomes great transformational success of all organization.

If you are dealing with other issues that hinder starting your robotization journey, please do not hesitate to share those. I’m sure we’ll find the way to overcome them. Robots are really your employees best friends !

Mariusz Pultyn
CTO

How robots can be a win-win in a typical company?

In all companies known to us one thing is taken for granted: IT resources are scarce and choices made in terms of what will be developed by technical team and what needs to wait for ‘unknown future’ are always difficult. Moreover, IT development is generally expensive what leads to very careful decision making with heavy emphasis on business case the investment is to generate.

IT resources are always limited. That means that choices regarding investment in technical development within companies are difficult. They result in many advances being left for improvement for ‘unknown future’. Also, IT solutions tend to be relatively expensive in implementing, hence natural direction is to invest in IT in areas which will result in creating new business and impact the financial results of a company more directly.

The above leads to a simple generalization, that usually IT resources are more likely to be placed into front-end, business development areas that are to have a larger impact on top rather than bottom line. Moreover, introducing new technologies, new channels, new business lines and new products very often leads to an increase in complexity in back-office. In ideal world this complexity should be hidden from our eyes and solved by automated interfaces and applications. As most of us know, in reality it is not always the case and very often more technology closer to front-end is associated with increase in manual work to be done in the back-office.

Hence having robots as automation mean in back-office allows for more allocation of more IT development resources towards areas with more sophisticated technology and potentially higher business growth impact.

Mariusz Pultyn
CTO

Robot manufacturing process automation

Is it possible to deal with large-scale robotization (more than a few robots) without automating the processes of creating, installing and monitoring robots?

We believe it isn’t. Robots used in RPA is a piece of software which sometimes needs to be changed. They work on applications which may also change from time to time what can cause even more complications. When starting robotization journey you should be always prepared for both of these factors and have tools to manage them in easy and safe way.

At the same time, we want the development of robots to be a simple and pleasant work, rather than a constant fight with the removal of errors. That is why in Digital Teammates we have prepared several tools and standards, that let us combine ease of working on robots (and thus shortening Time-To-Market) while maintaining a high level of quality and low vulnerability to errors.

Some of the main elements included in our solution for robot manufacturing process automation:

  • Standarization of the robot development process

Robots coming out of our factory are always characterized by certain common features regardless of which systems and in which environments they work on. This set of rules applies to both the manufacturing process (based on the best programming practices) and the construction itself (a set of rules that must be obeyed by all newly created robots). 

  • Creation of robots’ working environments

A very important factor, that can easily reduce the risk of possible errors is the standardization of robots’ working environments and the automatization of the process of preparing robots’ workstations. Thanks to this approach and dedicated tools, we are able to build virtual “robots’ desks” quickly significantly shortening Time-To-Market for our Clients.

  • Automatic tests, robots installation and configuration process

Our robots are tested automatically, and more specifically, are tested by other robots whose only purpose is to detect errors as soon as possible on test environments.

Fast and efficient installation of new robots also requires automation. Our robots use an integrated configuration management system, which helps minimizing risk of errors and  allows changes in start-up parameters without the need of installing new versions of robots.

  • Robots monitoring

All operations performed by robots are logged (on several levels) and stored in a dedicated log repository. The structure of these logs allows us to generate reports of both technical and business nature. Our clients have access to this data online and are able to monitor how do their virtual employees perform.

More details about mentioned tools and standards in my upcoming blog articles.

Bartosz Pietras
Head of IT

Hand in hand with robots – how people react to digital teammates?

As you can easily imagine, the decision to introduce robots that perform tasks handled by people can be a controversial topic for human employees. Our experience shows that even mere delivery of such news can stir the emotions and open up argumentative discussions. Both surprise and fear of the change itself and a potential loss of daily tasks can turn out to be a daunting blocker if not addressed in time.

Making employees accustomed to the subject and transparently presenting them what the planned changes are associated with is always a good start. Let’s not be afraid to say that robots will actually do some of our people’s work, but at the same time, we must not forget to emphasize that robots will handle the least liked tasks. Clear explanation of the fact that in the present reality we have much more work than resources to handle it is also a good argument because it’s our employees’ everyday reality and they can relate to.

A good introduction is one thing, but the best results in convincing people to robots stems from employees’ involvement in the very process of their development and the fastest possible launch of robots. These are the characteristics on which the Rent-a-Robot model offered by Digital Teammates is based. The most crucial here is the immediate transition from theory to practice what allows to convince our employees that robots are their allies, not their enemies. By handling the most repetitive and tedious tasks, they allow people to focus on a more creative side of their work and strongly emphasize the expert status of our human crew. Supervising the tasks carried out by the robot and at the same time necessity to handle most unique, complicated and complex cases by humans allows for the continuous progress of their specialist knowledge.

Reluctance related to employees’ anxiety of unknown and new technology is also easy to overcome by showing people that they are necessary for proper processing of tasks by the robots. The technological aspect and familiarizing people with the way that RPA works have an additional value, worth emphasizing in the conversation with our teams. Considering the dynamic growth of popularity of robotic solutions coupled with their relative novelty, first-hand experience gained by working side by side with robots, will be exceptional and almost exclusive for our employees. This is another aspect worth emphasizing.

In practice the introduction of robots can work wonders, especially in case of sudden spikes of volumes. Our employees noticed quickly that thanks to having digital teammates on board they are no longer asked to do overtime to handle additional volumes. The robots will take care of them gladly and without complaints and thanks to that our crew will be well-rested and much more effective.

Based on our experience, it can be noticed that despite the initial fears, anxiety of the unknown, and even resentment and controversy, robots can win the sympathy and favor of the employees. The combination of theory, elements of education and taming, and finally the launch of the first robots are enough for our teams to become convinced to RPA and to open them to accept the benefits of this solution.

To sum up the topic, I will quote the situation which happened few weeks ago – passing one of the employees in our office, instead of the usual “Hi!”, I’ve heard: “Magda, these robots are not so bad, after all. When will we get the next one? ” It seems that this short yet pithy point speaks for itself.

Magdalena Adamczewska
Head of Operations

How future workspace will change ?

Organizations we know and have seen are facing a challenge of having to few people for the amount of work they have to do. People available for work in back-office operations are scarce so it is a challenge to maintain the right team to keep the operations up and running. On top of this, many tasks that people are doing should have been automated long ago Their repetitive, tedious and error prone character means that these tasks should not be done by people any more and are perfect for robots.

We usually see that a typical team of 10 people has 7 more experienced and advanced specialists and 3 more junior members who are much more prone to attrition. This naturally leads to a situation where it is tempting to allocate the more tedious and repetitive tasks to the less experienced workers, what fuels the attrition even more.

Hence, having robots on board generally raises motivation as the most undesired jobs can be done by them and people, having Digital Teammates working shoulder by shoulder, can eventually focus on more creative and developing tasks what eventually leads to much more attractive workplace.

As we generally observe that typical robotization potential is ranging from 20% to 30% of processes, we can safely say that robots are perfectly fit to fill in the positions which are associated with highest rotation and the whole workplace immediately becomes way more attractive for the whole team, including the newcomers.

Konrad Jakubiec
Board Member

Do robots have cognitive skills?

Most of our robots strictly follow the procedure that is programmed up-front. This procedure sometimes can by very complex, because of the nature of the underlying business process. This allows the robot to work with 100% accuracy at all times. Robots can always explain why they made specific decisions and why they chose specific path based on specific situation. This makes them absolutely deterministic and perfectly fit for operations where the result has to be exactly, as provided by the procedure.

However in some situations we can benefit from artificial intelligence, especially when robots make some preliminary work, searching through big databases to prepare some data for further processing by humans. The examples of such activity from our experience can be reading incoming emails and flagging them with different categories. Even if the process is not 100% accurate it can save a lot of human work with reading all communication.

Mariusz Pultyn
CTO