Intelligent Automation in Enterprises
Albert Einstein once said, “If I were given one hour to save the planet, I would spend 59 minutes defining the problem and one minute resolving it”. Very wise words indeed, and when applied in the context of process agility and innovation with technology, enterprises would increasingly need to rigorously apply this to identifying and articulating the right problem to solve. Without that, they would end up missing opportunities while keeping busy needlessly investing in projects that are not aligned with their overall strategy.
Enterprises understand the tactical application of solutions like RPA. However, the true value add comes from taking this enterprise wide, that extends beyond point requirements. The insights that comes from automating processes is what drives optimal decision making.
So, what after all is Intelligent Automation? I asked a non- IT friend if he understood automation and Intelligent Automation differently? He answered when you do your automation process well it is called intelligent automation. I disagreed but I had to explain to him in a layman term and I attempted to use the example of a car since he was a car enthusiast.
My car has automatic gears and can sense my speed and change gears. That’s intelligent automation he said. Well, what happens when you see an ambulance flicking its siren behind you? I slow down and pull over. Who makes this decision I asked? Me obviously he said. Consider a driverless car that was trained, and more importantly have the sensors necessary, to differentiate between distinctly different scenarios, such as large, white vans and ambulances with red, flashing lights on the road. If the car is cruising down the highway and an ambulance flicks on its sirens, the car may not only slow down but pull over, because it perceives the ambulance as different from a big white van. That is intelligent Automation.
The same is understood in the business world as well. RPA is just the beginning of automation where we look at automating simple, standard, routine business processes not involving complex decision making. Enterprises can only unleash the value when it can meet customer needs and business outcome. Most business processes and decisions that enterprises work with, are not as linear and involves complex processes which needs to be done seamlessly between human and machines. In this setting, process automation requires cognitive abilities, such as natural language processing, speech recognition, computer vision technology, and machine learning to comprehend the vast amount of structured and unstructured data, learn on the go, and intelligently automate processes. The system will help make suggestions to improve processes just like how a human would. It helps enterprises to navigate through complicated situations improving speed, efficiency and decision making.
Business requirements are dynamic, and humans will always be an integral part of any automated process but what is critical is the close coordination and a hand-off process between the human and the robot that will make it intelligent. RPA works with standard processes that are rules based and is typically the first stage of automation leading to lower error rates and faster execution of the process. In a typical process requiring judgement and decision making, the process would be paused to complete this step before resuming. Where machine learning comes, structured, semi-structured and unstructured data could be leveraged to draw insights by enabling context sensing and decision making. AI chatbots could further automate the process when combined with robots to understand the question contextually and respond with robots.
Currently financial Services and retail are investing a lot on intelligent automation, but other industries will soon follow suit. For example, automating the underwriting process in line with the underwriter guidelines. Or when the machine can answer complex financial questions posed in simple English and the comprehension and presentation of the knowledge is done in the same manner an expert human does. The system personalizes the requirements based on the information collected and converses with the customer the way a human does including understanding the context and emotions of customers.
It is becoming increasing relevant especially for large enterprises who are mature in terms of their structured data applications and are now exploring the value they can derive from unstructured content and processes.
Here are some good practice guidelines on how enterprises can walk the path:
- Identify the right set of processes, define the problem and the desired business outcomes
- Processes which are critical and add tangible business outcomes if selected, will help reengineer the existing critical process and make it more simple and agile.
- It is important to identify the overall vision and the business outcome that is intended to be achieved. Working collaboratively across functions is critical for a clear gap analysis.
- Invest in the right automation software with cognitive abilities
- Build the right governance between human and Robots and clear hands-on and hands-off protocols
- Change management and adoption
The real question is whether one is looking for short term, immediate fixes to specific problems or more in terms of a new operational model that could be scaled enterprise wide and leveraged to meet business outcomes in the long run.
One of the first steps in selecting the right choice of vendor and implementation partner. Many enterprises evaluate several vendors and are sometimes influenced based on the sales pitches of various OEM’s. Sometimes this may lead to myopic decisions due to perceptions. The implementation should take care of aspects such as security, resilience and governance in addition to the automation aspects.
Intelligent Automation is effective when there are clear roles and responsibilities between human and robots. The decision of what will be done by human and what will be done by robot is a critical decision. Most intelligent automation fail due to lack of clarity on these protocols. The right approach to build the protocol is understanding clearly how a human executes the process and then defining what parts could be clearly handed over to a robot while building a strong monitoring mechanism. As business models evolve over time a change of one process although small could have wide spread implications across the business. Understanding the implication and re-training the human and robots is likely to be an imperative activity of the future.
Change management must be handled with diligence. The purpose, process and payoff must be clearly articulated and communicated to everyone. The biggest risk of not doing this is a false assumption that the robot will take care of business-critical decisions. Keeping a close eye on the bots and training them is paramount. Resistance and human tendency of blame game must be addressed as the bots and humans work hand in hand.
Intelligent Automation is still in its infancy; however, enterprises would immensely benefit and clearly have an early mover advantage if adopted correctly. For IA to be successful, enterprises will need to design these applications in a way that they are not siloed solutions that cannot be scaled. The design element needs to be given adequate attention to ensure they scale up to help enterprises achieve bigger business goals. As per Gartner’s prediction, “Through 2021, 40% of enterprises will have RPA buyer’s remorse due to misaligned, siloed usage and inability to scale.” This is not like buying software and implementing it, but it needs careful planning and design in such a way that the to-be processes being automated with RPA is carefully thought through before implementation.
IA therefore needs a more strategic approach to decision making and finalizing of vendors. Enterprises will demand greater proof prior to decision, that IA will indeed be successful in operating at scale. Articulating a problem and the intended outcome well could yield innovative solutions. In any case it is important to begin somewhere. Like Franklin D Roosevelt says “One thing is sure. We have to do something. We have to do the best we know how at the moment . . . ; If it doesn’t turn out right, we can modify it as we go along.”