I ask many clients the question – Do you think you need automation because you want to make your colleagues awesome or because you don’t trust them? And let me tell you – the ones that are focused on implementing automation because they want to make their employees empowered are the ones that succeed.
The road forward for automation within our industry is not clear and I would like to try to explain why we got where we are today and where I think we should move forward from here, specifically from within my area of domain: test automation. This area is of particular interest in the near future because the pure essence of test automation is all about how we provide decision support through the aid of machines.
Different ways to look on automation and the term ‘human factor’
Many people in the industry would argue against the definition above though. Test automation is a very popular focus area at the moment. But the general view in the industry is still stuck on the overall mindset to see automation as a replacer for tedious manual activities, and even more specifically – test automation as a machine that will identify and remove all bugs in our software before its released. The attached statistics below shows some interesting facts regarding the mindset on automation;
- almost half of the engineers globally consider themselves either afraid or confused with regards to automation whilst top management consider themselves neither afraid, nor confused.
- Sweden is falling behind when it comes to degree of automation implemented at scale.
From the report: “Upskilling your people for the age of the machine“
From the report: “Reshaping the Future: Unlocking Automation’s Untapped Value”
When looking at these interesting numbers, I get a feeling that engineers think automation is not their friend, but an enemy that could possibly take their job away. Another factor, when it comes to the Swedish aspect in particular, is the fact that in Sweden the phrase ‘the human factor was to blame’ in news reporting means that people did something wrong and is often to be seen as the cause of incidents where things go wrong. It then makes perfect sense that engineers find themselves reluctant to trust machines to do their job if the human part repeatedly seems to get the full blame when the real problem exists in the co-operation between man and machine. This is where it gets a bit complicated. How we perceive and look upon the nature of automation guides us on how we choose to implement it.
Why is decision support something important? How can it make us awesome?
The market is now slowly understanding that we are entering a new era where we leave the Tayloristic industrialization model behind, but what does this mean? When looking into the attached illustrated Taylor´s bathtub below and what it means – it tells us that the human has now the opportunity to become the most important part in the value-creating chain. During the industrialization, man was the supporter of the machine, now the relationship is flipped over. I believe that machines can support mankind to do great things. Machines are not to be seen as the bottleneck anymore because we can now create marvelous things with technology. However, this means that mankind is now in the driving seat and needs to steer, needs to know where we want to move forward, and here comes the complex part – The development process forward can never be seen as a straight line. We are moving into an era where source code is not our main offer to deliver, features are not valuable until proven to have a positive effect – development will become equal to ‘providing an experience’. And, when this movement is understood, and this mindset is established to its full extent, it will become very clear that no part of the development process can be seen as a linear process. It will instead be focusing totally on establishing an awesome experience for the audience of our provided services. The best possible way to develop this cannot be foreseen and must be explored in hyper-speed with the aid of machines and fast decision-making, done by humans with the support of intelligent machines. To read more into this subject I recommend you to download the ‘In Code we Trust’ industrial report.
From the report: “In Code we Trust”: www.sogeti.se/icwt-report
So developing software today is all about making decisions. We are constantly making small or big choices. Some might turn out to be right and some might to be wrong, but we constantly learn new things by continuously being on the move. Since the pace is going faster than ever, we need to accelerate our capacity to make smart decisions. In order to make decisions while running, we will become reliant on a complete new set of tools – hence everyone is shouting: we’ll need data, we’ll need analytics, we’ll need smart machines to help us, assist us. And yes, we will become dependent on artificial intelligence to support our natural intelligence to make us super-empowered. We need to transform ourselves to superheroes of decision makers throughout the development organization – with the important assistance from machines. But – the way to get there is a bit troublesome for some organizations. And it comes down to one question – are we willing to invest in a good experience for the employees within our development organization?
Why Employee experience is an important factor for automation
Why should organizations today focus on something like the employee experience? Isn’t it good enough if we just focus on the User eXperience (UX)? Yes, UX is the most important factor if you want to become successful in today’s disruptive market. But if you’re not paying any attention to your employee experience, or as I like to call it – the engineering experience – you will not be able to keep up with understanding and analyzing what your clients really need and appreciate. In addition to this – please also be aware that the new generation engineers are familiar and used to good UX from their personal experiences and don’t expect anything less than smart tools, great insights through data analytics and a high-velocity decision pace at work as well. I recommend everyone to read the book “Digital Workplace” by Oscar Berg and Henrik Gustavsson in order to dig into this area.
The analogy of deciding on what car to buy for your family
Since the area of implementing a good engineering experience is very abstract, I’ve found it useful to use analogies to explain. In this case – let’s look at the (developing) decision process of buying a new car for your family (your users). This is really a complex decision with a lot of factors to take into consideration. But if we oversimplify this decision, we’ll only look upon the act of making a car purchase. This means walking to our local car dealer, select a car and pay. If we aim to automate this process, we are most likely to build a robot that navigates to our car dealer and picks up a conversation with a sales person. This is natural to us since this is what constitutes a ‘car purchase’ for us. Next step would be to come up with the best agreement within the boundaries of what the robot and sales representative can agree upon. Then we perhaps say, ‘Hooray – this was a super-high-tech solution that we engineered and it was really complex to build. We are great.
Introducing the Model of the Five Senses in Intelligent Automation
But, in order to elaborate on why this setup is a bad initial automation investment to support our decision process, I’ll hereby like to introduce the model of Five senses of Intelligent Automation:
The model consist of five different senses that might be human-driven or automated to various degree independent of each other.
ACT – To perform an act, to complete a task or an action. This is basically where people or machines take the actual decision to make a move or react to some interesting event.
LISTEN/TALK – Interaction in any way with people and/or machines. This could be done by picking up natural conversations between humans, visualize data in a dashboard or use chatbots to interact between various parties.
WATCH – To monitor a system, to take in and register all its different aspects – from the full user experience to the technical monitoring of big amount of data sources from application, server, business and team data.
REMEMBER – Our capability to store data, our memory. This capability enables us to compare, to see historical trends.
THINK – Our ability to analyze, to create insights and advices based on what we know. To see complex patterns and trends in large amount of data forms the basis for our needed decision support.
This model offers us a way to think on automation in a more holistic way. Automation is not only meant to take away tedious activities and wasteful tasks in our mental industrial-work-assembly-line model. It could instead be empowering us, making us as human beings empowered. The basic idea is that it can reinforce us as humans enabling us to accelerate our capacity to make smarter decisions faster, and hence accelerate our throughput of real value. This model forces us away from thinking of test automation from an industrial perspective and makes is think of how to enable our employees.
To read more, please see: https://www.capgemini.com/the-five-senses-of-intelligent-automation/
Applying the five sense to our car purchase
Back to our analogy of our car-purchase decision process and put in the context of our five senses model. If we only focus our automation investments purely on the ACT of buying a car we might come to the conclusion that the smartest way to automate that activity is to program a robot to go down to the nearest car dealer with a TALK/LISTEN capability (chatbot program) and enough intelligence needed in order to make a successful purchase. However, if we instead aim to buy the best car possible to fit our family needs – considering ALL possible aspects – we soon will conclude that it might be more cost-efficient to start our automating efforts within the areas of WATCH, REMEMBER and THINK. Because, when you start considering various aspect of what car to pick we will have to understand our customer’s various needs when it comes to aspects of family size and economy, geographical location and so forth. It gets even more complex when we understand what car data attributes we need to gather and analyze, with regards to price, service data, tax details, mileage history and so forth. Once this data is captured (WATCH) we need to store this in a common repository (REMEMBER) in order to cross-reference all the data and look for different insights, trends and patterns (THINK)
As an example, look into the Car.info service online, it is a free service providing data and insights for more than seven million vehicles registered in Sweden. In this service a lot of information is available just by entering a license plate number and you will get tons of technical and financial information tied to that specific vehicle. Look at the enlarged graph from this service above, it only takes a second for a human being to understand that the specific car with this mileage trend graph is a bad decision to make. Once we as humans have understood that vehicles with this mileage trend profile is something we want to filter out from our list of possible decisions, we can learn our machines to exclude these automatically from our data sets in the future. We have then started to automate our THINK capabilities.
Start your automation journey today – from the right starting point
Once we have the WATCH, REMEMBER and THINK capabilities in place we have data to feed into the ACT and TALK/LISTEN capabilities and let me tell you one thing – if you start your automation road map from this way around you will most likely provide a super-empowered coworker with all the enablers needed to do a marvelous job. If an employee is asked to do a task, and while they do so, feel empowered by all the imaginable data accessible at their fingertips they are more likely to have a great engineering experience.
When looking back to what is happening in the software industry right now the pace is moving so fast that you are forced to totally change your mindset, and it starts now. As in the area of transportation – In the past, you were able to keep a car until it went retired. Now, people change cars normally after a couple of years (often due to leasing contracts ending). In the future, the culture of owning and leasing cars will most likely become an outdated consumer model and we will start having a transportation subscription where we’ll be able to pick the smartest vehicle to fit our needs in that moment. And in that future, your car-purchasing chatting robot will be recalled to as a bad investment.
So – why not start with automating the areas of WATCH, REMEMBER and THINK first, then explore further into the ACT and TALK/LISTEN areas. The companies that start implement this way around will stay competitive, will excel in engineering experience and co-worker attitude on automation.
So, my general recommendation to many clients right now on how they could tackle their automation investments are to form a step-wise automation road map where you start your thinking from your employees’ perspective and what they need to be empowered. And many land in the same overall road map outlined as below:
- WATCH-REMEMBER-THINK: Empower you co-workers with data, analytics, dashboards, visualizations
- ACT-TALK-LISTEN: Super-empower your co-workers by extending the reach of their capacity by adding robotics, process automation, artificial intelligence, machine learning.
But keep in mind – this suggested automation road map will never reach full automation in any of the five senses. Why, you might wonder? Guess what – the Industrial age is over – as long as humans are the dynamic part of the value creation, we will never be able to automate humans away from the equation. So, focus on your employees’ experiences and provide them with whatever enables them most. The illustration above illustrates a road of a healthy startup of an organization’s automation journey, but you need to find your way to enable the journey that suits you and your context the best to make your employees to get a good start – but where it ends, we can barely imagine.
http://www.dwstrategydesign.com/ the book “Digital Workplace” by Berg&Gustavsson