A mid-sized plastic injection manufacturer in Bursa decides to attach sensors to its production line. The general manager, impressed by an IoT demo at a trade fair, launches the project with a clear directive: monitor machine status in real time. Six months later, there is neither a reliable data stream nor a measurable improvement in production efficiency. The project stalls because the scope was never clearly defined and the pilot area was never properly chosen. This pattern repeats itself across IoT initiatives in Turkish manufacturing, and it points to a single root cause: starting with the technology rather than the problem.
IoT, the Internet of Things, refers to an architecture in which physical assets are connected through sensors and network infrastructure to collect data and support operational decision-making. The concept is compelling, but attempting to transform an entire factory or logistics network at once is a reliable way to sink a project before it delivers value. Successful IoT deployments share a common starting point: a well-defined, narrowly scoped pilot that produces measurable results within a reasonable timeframe. Pilot area selection is a strategic decision, not a technical one.
Three criteria consistently separate productive pilots from expensive experiments: visibility, measurability, and a realistic return timeline. Visibility means the pilot’s outcomes are noticeable both on the shop floor and at the management level. Measurability requires that success or failure can be expressed in numbers; vague assessments like ‘things improved generally’ make pilot evaluation meaningless. A realistic return timeline, in the context of Turkish SMEs, typically means three to six months. Pilots that stretch beyond this window tend to lose organizational momentum and face budget pressure before they can demonstrate value. The intersection of these three criteria points to where a pilot should begin.
In practice, the areas that deliver the fastest wins cluster around three categories: machine condition monitoring, energy consumption tracking, and inventory or asset location. In machine condition monitoring, sensors collect vibration, temperature, or operating hour data; because the cost of unplanned downtime is concrete and well-documented, the ROI calculation is straightforward. Energy consumption tracking delivers direct impact on the electricity bill within months, making the pilot’s value self-evident. Inventory and asset location monitoring, particularly in large warehouses or multi-site operations, makes search time and misplacement costs visible. What these three areas share is the availability of existing operational data for comparison; there is no need to construct a baseline from scratch.
One dimension that is routinely overlooked in pilot design is scalability architecture. Many companies treat the pilot as a standalone experiment, with no plan for what happens if it succeeds. This is a structural mistake. The pilot’s technical architecture should be designed as if it will eventually expand to cover a hundred machines or five facilities. Choosing standard communication protocols for the data collection layer, selecting a cloud infrastructure that can scale, and planning integration points that align with the existing ERP or production management system are decisions that must be made before the first sensor is installed. Treating the pilot as a temporary experiment means starting over from scratch when it succeeds, which significantly increases total cost of ownership (TCO).
The most common failure mode in IoT pilots is scope creep. A project that begins with ‘we will monitor three machines’ expands within weeks to cover quality control, logistics, and maintenance processes. When scope grows unchecked, measurability collapses; it becomes impossible to determine which variable is driving which outcome. The practical remedy is to define in writing, before the pilot starts, what the project will not measure. In Turkish IoT projects, it is this lack of managerial clarity rather than technical infrastructure problems that most often derails initiatives.
For a manager making an IoT pilot decision today, the practical starting point is this: identify the single operational process where you are losing the most money, where you understand the cause, but where you have no data-driven way to track it. That process is your pilot area. Define success criteria in numbers, set a six-month time boundary, and incorporate scaling architecture into the design from the first day. A visible, measurable, time-bounded pilot both builds organizational buy-in and generates the real-world business case needed to justify the next phase of investment.
This article was originally written in Turkish by Gökhan MERCANOĞLU on July 13, 2015 and has been automatically translated into English and other languages using machine translation.