Process-Driven, Data-Driven, or Model-Driven Automation?

The introduction of new technologies has enabled process automation to further develop beyond ‘screen scraping’ and data sorting. It has given rise to a new, smarter, and more powerful type of process automation: Intelligent Automation (IA). IA the automation of operational processes by using intelligent technologies such as Robotic Process Automation (RPA) and Artificial Intelligence (AI). It is used by many organizations today to automate more complex processes, freeing up the time of knowledge workers to be deployed on tasks that generate more value. These developments have created a wave of new terminology. Today we will discuss the difference between process-driven, data-driven, and model-driven automation.

Process-Driven Automation: Doing

Process-driven automation is simply driven by processes. Within this category of IA, the focus is mainly on performing tasks, via predetermined pathways and processes. It is not possible to deviate from this predetermined pathway as the process is the leading factor in this form of automation. When a task must be performed that does not fit within this predetermined pathway, and is therefore an exception to the rules, a human must intervene and take over.

The most basic level of IA is Robotic Desktop Automation. Basic tasks can be automated at this stage, often within one system and within one data silo. This is a virtual assistant to human workers, which can for example automatically send e-mails or generate reports. When moving a step up towards a more advanced level of IA, you will be introduced to Robotic Process Automation (RPA). RPA is great for automating slightly complex processes, sometimes across multiple systems. However, both are constrained to automating repetitive and mundane processes: doing the exact same thing over, and over, and over.

Process-Driven Automation is a great first step in digital development and transformation, however. It is manageable and can be implemented quickly, enabling organizations to start automating the first processes within weeks or months. This results in immediate efficiency gains, cost savings, time savings, and accuracy improvements. Process-Driven Automation is exceptionally useful when aiming for the ‘low hanging fruit’, by automating transactional processes: managing financial resources, streamlining customer services, and delivering physical products[1].

Data-Driven Automation: Thinking

Compared to process-driven automation, data-driven automation is a step up on the ladder of complexity in IA. In this case, automation is guided by data and context. This makes data-driven automation significantly more powerful and ready to handle more complex processes. Central to this type of IA is the combination of RPA with Artificial Intelligence (AI): transforming the intelligent robot of process-driven automation into the superhero that represents data-driven automation.

AI technologies includes, but is not limited to, Natural Language Processing (NLP), Machine Learning (ML), Deep Learning (DL), deductive and prescriptive analytics, and recommendation or decision-making engines. Good quality data is key, with AI being able to learn from this data and optimizing process automation as a result. This allows processes to be automated even faster and more accurately than with process-driven automation, across multiple systems and data silos. Unstructured data can be processed, and advanced AI can even perform human-like and judgement-based interactions.

The power of data-driven automation is the ability to automate non-repetitive processes, which significantly extends the range of processes that can be automated. Tasks that had to be performed by a human beforehand, such as processing scanned documents or telephone calls, can now be automated by AI technologies. In addition, RPA is becoming a lot smarter when it is data-driven, because IA enables processing beyond the predetermined pathway. However, all these opportunities and interesting possibilities also come with their drawbacks: data-driven automation can be quite complex, which drives up the costs.

Read more about RPA and AI here.

Model-Driven Automation: Doing and Thinking, Driven by Context

You may have noticed that we have not yet covered model-driven automation. Where does it fit into the landscape of IA? The answer is simple: model-driven automation fits  anywhere within the continuum of IA, as it is not so much about the degree of smart technologies and complexity used to automate processes. Instead, it is more about the way processes are captured within the process automation platform itself.

Figure: Unstructured data, through model driven automation, into applications.

The Be Informed Intelligent Automation Platform is an example of a model-driven automation platform. All knowledge about products, services, processes, and policies is captured in a model. These models are directly executable in an application, without the intervention of (extra) code, and can be easily interpreted by non-IT staff. This is often referred to as low code. The advantage of model-driven automation is that the models can be easily adapted if a change in the external environment occurs. This change occurs immediately in the application.

Using model-driven automation can involve varying degrees of RPA and AI, depending on the complexity of the process and the desired outcome. Therefore, model-driven automation can be used by any organization, regardless of its degree of digital development.

Get in touch with us!

We explained the differences between process-driven, data-driven, and model-driven automation. Are you curious about the possibilities of model-driven automation for your organization? Sign up for a free demo of our intelligent automation platform or get in touch with us. Stay tuned by signing up below or follow us on LinkedIn, Twitter and Facebook to stay in the loop at all times.

[1] IBM Institute for Business Value (2018): The Evolution of Process Automation