We have seen technology evolve from centralized computing in the mainframe era of the 70s to the decentralized computing in the client-server era of the 80s and 90s. With the advent of cloud computing, we are currently moving back to the concept of centralization. While we are yet to see full adoption of the cloud, this move toward centralization might be fleeting. I believe we're already moving back to decentralization, albeit in conjunction with the cloud. There are good reasons to believe this — big data is in distributed databases, blockchain is a distributed ledger, and perhaps most importantly, IoT is distributed computing.
Understanding IoT as Decentralized Computing
IoT is a game-changer, much in the same way as desktop computing and cloud computing were in their day, given its revolutionary potential, even more so. This time though, the delineation of centralized (cloud) and distributed (IoT) is rather blurry because they will both coexist and complement each other. While IoT is now a household term and needs no introduction, the idea of IoT as decentralized computing needs to be understood.
The term 'edge computing' (which might well become the next IT buzzword, just like ‘data science’ was previously) is widely used to explain the function of IoT and the future of technology in general. Edge computing means that the 'things' (of the Internet of Things) will not just act as sensors and actuators of a microcontroller. They are moderately powerful computing devices that can process within the device without communicating with the internet (or the cloud) for improved performance. The device will be at the logical periphery (hence the name 'edge') of the system of devices and gateways.
This has been made possible due to a massive reduction in the cost and size of memory and processors. Even though there is a lot more you can read on why it's important and how it is implemented, the primary focus of this article is to explore the possibilities this brings to the world of Enterprise Resource Planning (ERP) and if we are moving toward an era of distributed ERP.
Are We Heading Towards an Era of Distributed ERP?
The use cases for IoT in the context of ERP are many. For example, it can be used for product improvement by providing real-time feedback. It can enhance warranty services and warranty management. It can be used for predictive or preventive plant maintenance. The list is endless. However, all of these operate within the restricted paradigm of the current idea of IoT, where the device acts more of a sensor than an actuator.
However, IoT takes on a whole new meaning if you think of it as a part of a distributed ERP system. A distributed ERP means that the data is located not just on a centralized server (either physical or cloud), but also distributed across devices and systems. This is an example where cloud and edge computing will coexist and complement each other. So, the data which is relevant to the peripheral device, either generated in the device or user-generated and synchronized from another location, will be present within the device itself. This results in improved latency and better utilization of available resources.
With the concept of distributed ERP, several new possibilities and use cases emerge:
- In cases of predictive plant maintenance, the equipment acts as an IoT device, and the historical data of the equipment will invariably be present on it. However, if we can also store sales-related data and schedule lines in it, the equipment can fine-tune itself with a little help from machine learning techniques. Therefore, the promised delivery times are not compromised even in case of preventive maintenance, for example, by shutting itself down at suitable intervals.
- In manufacturing, a production line acting as an IoT device can feed the actual production data for a batch directly to the ERP database without user intervention (or without using a tool enabled by Robotic Process Automation or RPA). Now, if this can be combined with data from the Quality Management (QM) module (and the sales and distribution, or SD, module, in case of Make-To-Order or MTO production), the actual input to the production line can be calibrated to produce the desired product.
A Distributed ERP Ecosystem is Closer Than You Think
In a distributed ERP environment, each IoT device will be able to collect data from other modules and process information within its processor, creating a smarter and faster device. As devices become increasingly connected, and user data maintenance reduces, the ability to combine the data generation with information from related modules will be instrumental in downstream reporting, decision-making, and preventive maintenance.
Despite rapid advancement, IoT is still in a nascent stage, and the distributed ERP environment as a concept has not yet been conceived. This article only explores the possibility of enterprise technology moving in this direction. But with all major vendors, including SAP, investing heavily in these technologies (SAP, with its investment in SAP Leonardo), we can foresee an ERP ecosystem shaping in this direction sometime in the next decade.