BACKGROUND IMAGE: iSTOCK/GETTY IMAGES
A solid business process management foundation will help businesses keep pace in using robotics, but enterprise...
architects must adjust to more chaotic flows to take full advantage of it. Here we examine the challenges of rolling out robotic process automation, especially change management and organizational costs, and how organizations can tackle those challenges head-on.
Robotics is quickly moving out of the realm of science fiction into business fact. Businesses like Uber and Airbnb are deploying disruptive models, which can be augmented by robotics. BPM technologies must evolve to support this transition.
“BPM is going to be a critical leverage point for the enterprise deployment of robotics,” Nathanial Palmer, VP and CTO of Business Process Management Inc., a BPM consultancy, said.
The biggest challenge in rolling out robotics is the change in management and organizational costs. BPM will make it possible for the enterprise to leverage its existing infrastructure and security models in the emerging era of robotic process automation. Palmer said that BPM is fundamentally a discipline involving any combination of modeling, automation, execution, measurement and optimization of business activity flow in support of enterprise goals, spanning systems, employees, customers and partners within and beyond the enterprise.
The move towards improved automation requires appreciating the growth of robotics, Internet of Things (IoT) and the use of mobile technology. Enterprise architects need to plan for chaotic workflows using the torrent of information from these devices. Leveraging process discovery tools can help enterprise architects improve the overall business rather than specific processes. A transition to 4-tier architectures can bring agility to implementing these improved workflows.
Create guardrails rather than conveyor belts
Today, business processes move through well-established pathways, like conveyor belts moving goods. In this existing model, the conveyor belts of business processes don’t care about the contents of the boxes being moved.
This approach works well in traditional workflows when the process is well understood. Languages like BPMN make it easy to characterize a well understood process for ensuring compliance to ensure that all of the appropriate steps are included in a process. This approach allows consistency for specifying a process, communicating it between managers and workers, and creating executable business processes.
To keep pace with disruptive business models, enterprise architects will need to think about creating guardrails for more chaotic processes. For example, the Amazon Robotics warehouse leverages a fleet of pallet robots to autonomously move goods around the warehouse. Each robot can move across many different routes between stored goods and the shipping station in the fulfillment of an order. Since these robots cross paths, a longer trip for one robot sometimes means a much shorter trip for others.
Inspecting the contents of processes
This new style of more dynamic BPM has to understand the payload inside the boxes and the whole chain of events related to the contents to make better decisions. This applies to the movement of physical goods and the information relevant to a business process. This new style of BPM is driven by goals and can help to expand the envelope of automation. The practice of BPM in and of itself is great at automating a process. But this needs to be intelligent automation, or the enterprise can end up optimizing the wrong process.
Three important components driving more intelligent BPM include the rise of robotics and smart things, rules and decision making, and the use of relationship data. The growth of IoT is making more information available to understand the complex relationships the enterprise is connecting to.
Prepare for new experiences
Robotic process automation is not just affecting the construction and logistics of consumer goods. For example, the banking industry is now starting to take advantage of a new generation of social robots to support new user experiences.
“The state of the art is how to shift from human and capital intensive branches to ones that rely more on automation," Palmer said. Bank of America was one of the first major banks to deploy document scanners that provide many of the services of human tellers like loan origination and account opening. In China, banks are also deploying more human-like robots in some branches to augment human staff.
At the same time, banks want to provide as many services as possible via new consumer channels in addition to mobile phones, and PCs, like the Amazon Echo audio-only interface or smart watches. BPM can provide the underlying framework for supporting these new channels. This requires allowing the systems to react to the payload. Adaptive systems can change course in response to an evolving operational context by leveraging rules and context to expand the envelope of what can be automated.
Adopt a 4-tier architecture
Palmer advised enterprise architects to think about moving beyond the 3-tier architectures that are common today. This requires developing a 4-tier architecture that consists of a client tier, delivery tier, aggregation tier and services tier. The client and delivery tiers are separated apart in this new architecture to allow data and application capabilities to be rolled out to new clients quickly.
That data tier and aggregation tier replace more tightly coupled service oriented architecture (SOA).
“The systems have to be able to be decoupled from how work is conducted, and that is how we will integrate robotics with BPM," Palmer said. "Every event has a half-life. Leveraging BPM and robotics can shift the point at which business events come actionable. The business case for many aspects of how robotics will be deployed in the organization is that shifting that point forward. ”
Robotics make more information available for decisions
A side benefit of robots is that they can gather more information about events than humans. Today robotic tellers are a bit of a novelty, but as the sensors and tools for sentiment increase, they have the potential to gather far more data about a process than is practical for human tellers. Palmer said that the enterprise can mine this additional data to understand the best process they could do for a particular type of transaction or as part of the sales process of relevant new products.
Most enterprises do process mining at the level of transaction systems. But, according to Palmer, this only captures about 2% of what actually occurs in the business. The wealth of data generated by IoT sensors, robotics and smart phones could enable businesses a much better understanding of what occurs during a process. A new generation of process mining tools will make it easier to shape the direction of business based on this data coming in real time. Palmer said some of these leading process management tools include offerings from XMPro, Whitestein Technologies, Appian, SAP and IBM.
“We started with the notion that not all processes are linear," Peter van Schalkwyk, CEO of XMPro, said. "About 80% of processes are unstructured. As a result enterprise architects need to look at emergent processes. “
For example, in a sales process, the organization might notice that one salesperson consistently outperforms the others. This might just be the result of the order they present the various steps to the customer.
“Sometimes longer paths yield a higher revenue," Schalkwyk, said. "These are the kinds of things you look at with automated business process discovery.”
Weaving BPM into your social media processes.
Building microservices for BPM: The next big thing?
Creating flexibility with agile BPM tools.