Consider a mid-sized automotive parts supplier in Bursa: sensors on the production line, an ERP system running daily operations, and e-invoice exchange already in place with key suppliers. Yet these three systems do not talk to each other. Sensor data stays on the production engineer’s desktop. Inventory figures in the ERP get updated manually once a day. Supplier delivery timelines are still tracked by phone and email. The problem is not a lack of technology — it is that the technology in place was never designed to feed into a coherent operational whole.
Big data, the Internet of Things and smart manufacturing have been appearing with increasing frequency on the agendas of Turkish industry conferences and trade publications over the past two years. They tend to be treated as separate topics: one belongs to the IT department, another to the production floor, the third to the general manager’s strategic agenda. That separation is artificial. Big data is the analytics layer that processes the raw material IoT generates. IoT is the sensor network that captures real-time signals from production processes. Smart manufacturing is the operational framework that converts both into decisions. Building any one of these without the others is like buying a vehicle without thinking about the fuel infrastructure.
On the IoT side, machine sensors are not new to Turkish manufacturing. Temperature monitoring, pressure measurement and fault detection have been in use for years. The real issue is that the data these sensors produce has historically been stored locally, in isolation. With the strengthening of broadband infrastructure and the maturation of industrial networking protocols, moving that data to a central platform is now technically feasible. In Turkey, the first meaningful deployments are happening at larger industrial firms; SMEs are typically running pilots on a single line or a single process, which is a reasonable starting point.
The big data side is more complex. The term is often associated with petabyte-scale datasets, which leads many SME managers to conclude it has nothing to do with them. That conclusion is wrong. Operational big data makes sense at far more modest scales: combining hourly sensor readings from a production line with supplier delivery variance records and customer return patterns already produces actionable insight for a small manufacturer. The critical concept here is the capacity to handle unstructured and heterogeneous data — not the summary tables that a standard ERP reporting module generates, but the ability to store and query data from multiple sources in a way that supports pattern recognition.
The operational payoff from connecting these two layers — IoT infrastructure and data analytics — is most visible in supply chain management. Take a textile manufacturer: if a yarn supplier’s production-side sensor data can be integrated with the buyer’s ERP material planning module, a supplier-side disruption risk can be identified before it affects an order delivery date. This scenario is being piloted at larger firms today. The infrastructure it requires — ERP integration at the data transfer level, a reporting layer, basic alerting logic — is also achievable at SME scale. The cost and timeline differ, but there is no structural barrier.
For the executive, the central decision is whether to treat these three waves as separate budget line items or as a single operational maturity program. The second approach produces a lower total cost of ownership (TCO) and reduces implementation risk. When projects run in isolation, each layer accumulates its own technical debt. When the data architecture is designed from the start to accommodate all three layers, subsequent steps require far less rework. The ROI calculation also becomes more honest: the return on a sensor investment cannot be measured without the analytics layer that makes use of the sensor data.
Three questions determine whether an SME is ready to move. First, does your current ERP accept external data — even at the level of file import or a database connection? Second, can you name at least three critical parameters on your production floor that you want to monitor but currently cannot? Third, what is the single largest source of uncertainty in your supply chain — supplier delays, quality variance or demand fluctuation? A firm that can answer these three questions clearly already knows where to focus an IoT, data analytics and smart manufacturing investment. The technology roadmap should start from those answers, not from a vendor presentation.
This article was originally written in Turkish by Gökhan MERCANOĞLU on July 22, 2013 and has been automatically translated into English and other languages using machine translation.