Headless “Digital Core”
I find it a strange irony, that just as SAP have sorted out the way users interact with SAP (see the Fiori Design Guidelines) a user interface starts to become superfluous to requirement!
In the same way that warehouses have been transformed by automated picking machines, I am starting to see a future where the need for a user interface via a screen / device will become the exception and not the rule for the digital core. This is being driven by the following advances in technology:-
- Real-time integration
- Cheap Sensors
- Predictive Algorithms
- Machine Learning
- Voice Recognition / Pervasive Mobility
I think that bringing all of these capabilities to the digital core will mean that the day to day “logon to SAP” will become redundant. Instead, people will respond to alerts about specific business situations that need to be addressed, which will keep business processes running as smoothly as possible. If we can have self-driving cars, why not a self-driving Digital Core?
Perhaps if you do have an SAP logon in future, it will show you something like this:-
So how could that work??
The key foundation for this automated “Digital Core” is the need for real-time integration of data and transactions, so that the real world and the digital world are as aligned as possible and we remove the need for anyone to “key transactions” into SAP. This requires 3 major building blocks:-
- High quality data
- Well defined mapping between the source and target systems
- A robust and fault tolerant integration landscape
Only with these in place can we start to build smart solutions that react to what is happening in real time. You wouldn’t want a self driving car that based decisions on data that was 5 minutes old!
Sensors on Everything
As the cost of sensors / tags goes down, it becomes possible to more closely create a “digital twin” of what is happening in the real world.
It could be associated with physical goods (containers / totes etc), but could also be applied to information you could automatically gather from your sales force, in terms of visits / e-mail etc, for engineers recording visits / parts consumed etc. and increasingly from your customers, based on their location and activity for example. When all of this data, (with the appropriate security checks and balances) is integrated into your systems, you can give your automated algorithms the best chance of making the next best action when anomalies are detected.
So now we have lots of data of high quality and in real time, this is the secret sauce that makes predictive analytics fly. With the right predictions we can start to look into the future and identify potential problems before they become real problems. From here we can alert the right people of potential problems or hand off this knowledge to automated compensating processes (e.g. expedite purchase orders for materials where we predict shortages).
Things get even more interesting if we start to hand these business situations to algorithms that are capable of learning what types of decisions are preferred and which decisions create the best outcomes. For this we enter the world of machine learning where computers can take over “fuzzy” tasks that could only be handled by people in the past (e.g. Fuzzy Cash Matching).
Voice Recognition / Pervasive Mobility
And when things get too much for the machine learning algorithms, they can always “phone a friend” and have a conversation / text (or Instagram / Snapchat for younger users), enabling real-time feedback into the process. This feedback can also be used to “train” the machine learning components to make the decision automatically in the future.
If this all seems a bit too far in the future for you as you stare at SAPGUI – I suggest you start to look at S/4HANA, as all of the above features are being baked into the solution – welcome to next generation automated ERP.