We’re all familiar with the image of robots in manufacturing. These large industrial machines are purpose built to automate highly repetitive, and potentially, dangerous tasks. The fact is, these robots are the result of over 100 years of innovation in the manufacturing process which we typically call the “Industrial Revolution”.
It seems business professionals are still seeking this “holy grail” of automation achieved in the manufacturing industry. There are many tools being built that claim to accomplish these goals, yet as you dig deeper, you quickly realize these systems are no more than basic tools that humans are utilizing to manually operate a process.
To illustrate this point, suppose we want to manufacture book shelves. A screw driver is a tool that makes it a lot easier to fasten wood. We could hire a bunch of people, give them a screw driver and a box of screws and have them build the shelves. Back at the beginning of the industrial age, that’s exactly what we did. However, not only is it inefficient – it also leads to inconsistent results and, most likely, a lot of repetitive stress injuries.
Let’s say we upgrade those screw drivers to electric drills. Now the human becomes a lot more productive and less likely to be injured. However, we still have inconsistencies in the construction of our shelves since different workers may use more or less screws, fasten the wood in different spots, and not always connect pieces well enough.
So we create rules that every worker needs to follow to build a shelf. That fixes the problem to some extent, but introduces several other issues. First, there’s a lengthy training program required to be proficient in shelf building. Second, we introduce inspector workers to the process to assure everyone is following the rules to build shelves according to our standards. Clearly, our shelf building operation is going to have problems scaling up since we now need a lot more people to produce just one shelf.
Suppose instead of adding a layer of human operated rules and oversight to our operation, we develop a more automated mechanism to both improve the effectiveness of our human shelf builder and apply our rules to properly constructing a shelf. We invent a device that a human loads the parts of the shelf to assemble. This device has drills attached in the correct locations for all the screws. Once loaded, the human operator pulls a lever and the device aligns all the wood, loads the screws, and drives them into the wood.
This form of “robot” is exactly what you see today in manufacturing. Devices are purpose built to aide a human operator to quickly construct goods in an efficient and consistent manner. Humans are utilized according to their skills of intuition, adaptability, and reasoning while robots manage the repetitive nature of a job.
If manufacturing has done so much to adopt robotic automation, why are business professionals still handed a screw driver and box of screws and asked to build a “shelf”?
Of course, by screw driver, I mean various software tools like email, spreadsheets, and record management systems (CRM, HRMS, etc). And the “shelf” is whatever service your providing within your organization. All of these are great tools but generally require a great deal of human effort to produce an actual product.
Now, I’m not saying we need a physical robot to help automate a process. Instead, we need more purpose built “software” robots to help align these tools in a way to enable humans to use their “human” skills to produce a consistent product in an efficient manner.
Before we can deploy our fleet of software robots, we need to determine what they’re actually capable of doing and how those skills can be interwoven into our business processes. Since so much of the business world is run on data and not physical items, it can sometimes be hard to visualize areas in a process where a robot can help. However, if we look at data as a package that needs to be sent somewhere, opened, modified, repackaged, and sent to another destination, it becomes more clear how software robots can be introduced to manage certain aspects of that process.
At this point, an example will help clarify these concepts better. Since I’ve spent most of my career around the staffing industry, I’ll use the process of a recruiter placing a candidate and initiating the onboarding process. I’ve found the process of handing off the placed candidate to the parties responsible for managing the onboarding process to be quite tedious as a large amount of information needs to be transferred to ensure all the knowledge of the new hire is disseminated to the correct people involved in the process. Many times, email is the tool of choice since its already deployed and everyone already uses it.
Let’s pretend our recruiter sends a email with details about the new hire. We may have told all of our recruiters that they have to include specific key information in the email so all the other people receiving it know what they need to do next. However, emails start out blank and things get forgotten so our recruiter sends incomplete information. Since different people need different information, they all start responding asking for the bits they need to move along. After 27 rounds of emailing, all the details are worked out and onboarding can start.
Now, let’s build a software robot to make this process less chaotic and produce better, more predictable results. Instead of emailing everyone involved in onboarding, the recruiter emails the robot with whatever details he has about the placement. Using artificial intelligence to recognize and extract the details of the placement, the robot prefills a placement form and requests the recruiter to supply any missing details. Once complete, the robot initiates the onboarding process by sending out paperwork and providing access to the people involved in the process.
The software robot essentially is orchestrating the process and ensuring all parties have the information they need for performing the next steps in the process. Email tools can still be used since they provide a natural interface to starting the process. However, without someone or something “policing” the process to ensure order, it is quickly crippled by its inherent weaknesses. Try to scale this up to multiple placements from multiple recruiters, and the whole process begins to breakdown.
Notice in the example that we introduced a record keeping tool into the mix. However, instead of forcing humans to do all the steps to start a record and rekey information from an email, the software robot connects the tools together in a specific why to eliminate unnecessary steps and move the process along.
There’s an inherent fear that software robots will eventually replace all of the humans leaving a large population of people jobless. However, I have a more optimistic view that these robots will make humans more productive and derive more enjoyment from work. Instead of the grinding repetitive tasks we do today, software robots will allow us to focus on the creative nature of our jobs like human interaction, research and development, data analysis, and decision-making. Demand is increasing for these “knowledge products” but our production is being stifled by the growing amount of manual labor being introduced into business. The human-race created technology to serve us – its time to innovate our business processes and deploy software robots to do our bidding.