Big data has certainly been the hottest technology trend of 2016, with many companies frantically rushing to implement different big data projects, worried of falling behind competitors. The natural progression of data collection is data and process automation, and for organizations serious about making big data gains, now is the time to take this logical step.
Decisions are made by analyzing data, drawing conclusions from it, and ultimately executing upon those conclusions. The massive rise of data acquisition and analysis has enabled us to draw amazingly complex and intricate conclusions, and automation is providing the best way to facilitate this process.
Considering the enormous growth of data sources available to many organizations, Enterprise Process Automation (EPA) using software robotics has never been more accessible or necessary to organizations. However, while many executives get excited about the fact that they can automate more than 80% of entire business processes end-to-end, we need to remember that a robot is only as good as the data it’s fed.
Robots – What Are They, and How Do They Work?
When thinking of automation software, many think of simple scripts designed to complete one specific, repeatable task. However, the enormous access to data we now enjoy has given rise to a new phenomenon – software robotics. Instead of focusing on a single task at a time using swivel chair automation technology at the user interface, a software robot can manage entire business functions such as the supply chain or the internal finance processes of an organization. Robots not only automate each individual job within these operations, but manage and control the interdependencies of each section as part of the wider ecosystem. Managing the entire ecosystem allows an organization to enjoy greatly increased efficiency as well as provides a much higher level of quality control, reduction of errors, and a complete and accurate audit trail.
As you can imagine, this requires just as much if not more data that a human would require if they were to be doing each part separately. This is where the importance of good data management comes into play, and cannot be understated.
The Role of Data, And How It Is Collected
It’s important to understand the different data flows available to software robots. Good robotics solutions are data agnostic in the sense that they can take any form of data and convert it to a standard format which the robot can then use. As such, the best source of data is establishing a direct connection with a data-generating application, such as an Oracle system, to guarantee real-time and accurate updates. If this is not available, a middle layer such as SAP is the next-best data reservoir for a robotics solution to tap.
If an organization does not use a higher level data processing tool, the robot can sync directly into most data lakes, or even just a simple SQL database. However, the higher the level of synchronization, the lower the chance of entry miscommunication or other errors when the robot is performing various tasks.
Data is collected through a relatively simple and straightforward method. The robot requests the data, which then prompts a user to verify whether it can be collected. Next, the data is retrieved and converted to a simple HTML or flat file within the system, although this is not restricted to any one output format. Once the data has been structured and standardized in the robot’s systems, it can begin to derive conclusions from it.
While the end result of robotic data collection is format independent, a lack of standardization of data storage has the potential to cause difficulties, or errors in data retrieval. While the obvious solution involves standardizing an entire organization’s data into one structure, most data engineers know this is simply not feasible. Different departments use different software, and many companies use third parties to gather and hold their data. In this case, it is important that the robot consultants have a clear map of all data sources within an organization, and how each segment is structured in order to program the robot to read each source independently of one another.
Structure, Structure, Structure
This raises the most vital point for robotic deployments – developing a clear structure and goal before an organization begins. As with any data project, executives can sometimes focus too heavily on automating for automation’s sake, and lose track of the wider objectives. This will not only defeat the original business purpose, but create difficulty for the data engineers responsible for organizing and maintaining databases, as they will lack clear structure and direction.
Going hand-in-hand with the hype around big data, artificial intelligence (AI) has been a topic of much industry speculation and excitement, with many seeing this as unattainable, or wishful thinking. However, AI is really just a mathematical application of information in the same vein as the human decision-making process – review, process, analyze, draw conclusions, and finally execute upon data.
Devin Gharibian-Saki is the Chief Solution Officer at Redwood Software.
Calling on 10 years of experience in SAP, having worked as a technology consultant, project manager and process advisor, Devin brings a wealth of process automation experience to Redwood that he brings to bear daily on both internal and external audiences.
In his capacity as CSO at Redwood, Devin is responsible for ensuring that their Robotic Solutions meet the actual customer demand in the Robotised Enterprise™ and, when delivered, that these solutions are implemented with the highest quality either solely through Redwood, or in collaboration with partners.
Calling on his diploma in Mathematics and understanding of the typical challenges that large organizations are facing in managing their processes, Devin solves complex problems by developing the appropriate process automation strategies using a combination of SAP ERP and technology skills, Process Optimization proficiencies and knowledge of robotization technologies.
He also holds two patents for good measure.
Subscribe to Data Informed for the latest information and news on big data and analytics for the enterprise.