IoT: 7 ingredients to design modular & scalable device
- Vinay Solanki
- May 17, 2025
- 4 min read
The Internet of Things (IoT) is rapidly expanding as a technology, influencing various sectors including manufacturing, retail, automotive, infrastructure, logistics, public services, and consumer markets, among others. It is expected to create billions of dollars in value, according to research organizations like Gartner, McKinsey, and IDC, as well as industry bodies and experts, due to its extensive applicability in both B2B and B2C contexts. For the average person, IoT can be described as a data enabler that uses sensors to gather information, providing valuable insights and benefits to the user.

The IoT device market size is anticipated to range between $300 and $500 billion USD by 2025, making it a highly attractive value proposition. Connected devices include sensors, actuators, connectivity, processors, peripherals, and passive materials. For instance, a temperature and humidity sensor in a truck transporting perishable goods tracks the readings and uploads them to a cloud or remote server in real time using cellular connectivity (4G). The data is then processed to generate alerts or take necessary corrective actions.
In certain cases, processing can occur locally within the device's processing unit to save time, reduce bandwidth usage, or generate critical alerts. These capabilities contribute to edge intelligence in the solution, also known as edge compute, as illustrated in the image.
Typically sensors are inside the end devices such as wind mill, factory equipment, connected car or retail shopping trolley and these sensors communicate with gateway or smart hub that do more complex processing and enable long range internet connectivity with the cloud.

IoT device manufacturing ecosystem comprise of participants including but not limited to OEM, ODM, Embedded Designers, Firmware Developers, Hardware Architects, PCB Designers, EMS service providers, Mechanical Designers, Testing & Validation labs, Certification bodies and so on.
IoT device design need to be highly reliable, modular, secured and economical with the capability and capacity to both prototype and mass manufacture. Device productization with a scalable first mind-set is the key and here are 7 steps to consider:
1. Ideation: Detailed requirement analysis of the use case is key to design an optimal architecture and firmware logic with consideration for price, form factor, security, interfaces, data transfer support, edge analytics among few other. Ideation also help to analyse if off-the-shelf or a readymade device may fit the requirement and if matches a minimum of 80% then it is advised to customize available design.
2. Hardware Design: The next step involves designing the electronics and mechanical components. The electronics design begins with a block diagram of the PCB, detailing all interfaces, modules, components, and their interaction with the application layer, such as a 4G module, Wi-Fi chip, battery, antenna, and memory for a basic 4G-based tracker. The hardware designer then creates a detailed architecture diagram, focusing on modularity to ensure the design is not tightly bound to a specific component manufacturer, but rather remains as agnostic as possible to the chip or module provider. It's advisable to keep the design flexible enough to accommodate changes, aiming for an 80% fit with the expectation of 20% customization in IoT. For instance, if a customer requires Wi-Fi and BLE, instead of opting for a Wi-Fi-only chip, if the budget allows, choose the ESP32 that supports both, while keeping the architecture adaptable to support both ESP12 (Wi-Fi only) and ESP32. The mechanical designer collaborates closely with the electronic designers to ensure that the enclosures and other mechanical design aspects meet product requirements, such as achieving an IP67 rating.
3. Firmware Design: Selection of programming logic to support the use case goes hand in hand with the hardware design hence the consideration should be for a reliable low power embedded architecture that can run on the hardware and fits memory and power requirement. Data security and encryption (ex: AES 256) is crucial design driver to protect data stored on the device or transported from the device to the network. Choosing hardware with a good developer community and resources will help to develop firmware in time.
4. Prototyping: Building MVP prototype with the electronics and firmware to test functionality and performance requirements with possibly off the shelf or 3D printed enclosures without investing into tooling/moulding cost upfront will reduce sunk cost. Prototype can be created at lower cost and given to customer for testing and validation. However DFM (Design For Manufacturing) processes should be adopted to ensure the final design is manufacturing friendly and can help to keep the cost per unit low.
5. Value Engineering: Post sample approval you can optimize the BOM (bill of material) to ensure we do not compromise on quality of electronics or mechanical components but bring price down to stay competitive. Volume, component type and power of negotiation all help to add value i.e. value engineering or analysis can lead to customer satisfaction and repeat business.
6. Testing & Certifications: Post finalization of the BOM and samples, we move to final UAT and validation to ensure product in its close to final form supports all functional and non-functional parameters and KPI. If any certifications, which could be costly and time consuming such as FCC, CE, BIS, ATEX, IATA, FAA, etc. should be taken up at this step.
7. Mass Manufacturing: Finally when it comes to mass manufacturing it is key to optimize lead time and process cost in order to reduce production price per unit assuming a reasonable volume. Driver of a uninterrupted product is availability of raw material and hence strong supply chain management of all input material (electronic, mechanical, consumables, etc.) with proper inventory management must be in place. Processes & machine settings on the SMT and assembly lines need to be well defined and must be in line with traceability solution to capture the raw material information. Additionally testing details, packaging, documentation for quality assurance and shipping information should be handy and ready.
Constructing a modular device is more challenging than it appears, so the chosen approach should be based on a thorough understanding of market needs, product use cases, and the diversity of customer requirements. Sometimes, having a product that is 80% complete with one or two variants available for demonstration, along with the ability to customize the design, is acceptable.

Comments