How to Choose the Right Hardware Platform for Connected Products

Cover Ble 09 19 2025@4x 8

Choosing a hardware platform is one of the most consequential decisions in connected product development. It shapes everything from your power strategy and connectivity stack to cloud integration, security posture, and development cost.

This guide breaks down five foundational decisions that shape hardware success for connected products, grounded in real-world lessons from helping teams bring Bluetooth-enabled and cloud-connected devices to market:

  • Wireless connectivity: What communication model actually fits your deployment needs?
  • Embedded OS: Should you stick with RTOS or invest in Linux?
  • Cloud protocols: Will your hardware and platform work together or block each other?
  • Vendor selection: How do you balance support, availability, and future flexibility?
  • Integration readiness: What will accelerate, or stall, your time to market?

There’s no single right answer, but there are patterns. Knowing which trade-offs matter and how to prioritize based on your product goals is what makes a decision scalable. Whether you’re building from scratch or optimizing what you’ve already got, this article will help you make hardware decisions that support long-term performance across the full system.


Examining Current Wireless Technology 

The wireless protocol you choose has far-reaching implications that shape your product’s cost structure, power profile, system complexity, and long-term viability. That’s why it’s critical to evaluate your current wireless setup against your product’s actual deployment needs. From communication range and bandwidth to update strategy and security model, this decision affects every layer of your connected system across the product lifecycle.

While the wireless landscape is broad, most connected products share a common starting point. BLE and Wi-Fi cover the majority of practical short-range needs, offering a strong trade-off between performance, power, and complexity. Other options like LPWAN, Zigbee, and cellular are best reserved for more specialized deployment environments or functional constraints.

Still, not every product fits the default. If you’re designing for extended range, ultra-low power, or deployment in rugged or infrastructure-light environments, one of the following may be a better fit:

  • Alternative short-range protocols (e.g., Zigbee, Thread, Z-Wave)
  • Low-Power Wide Area Networks (e.g., LoRa, Sigfox, NB-IoT)
  • Traditional cellular technologies (e.g., LTE-M, 4G/5G)

The diagram below maps these technologies by range, data rate, and power consumption, highlighting where each excels within the broader IoT landscape.

A scatter plot mapping wireless technologies (BLE, Zigbee, Wi‑Fi, LTE‑M, LoRa, etc.) by data rate and range, with color coding for power usage. Shows how options vary from short‑range, low‑power solutions to high‑range cellular networks.

Choosing the Right Wireless Technology

With so many options on the table, it’s easy to get distracted by specs or feature sets that don’t align with your product goals. A structured approach, grounded in communication needs and deployment context, helps cut through the noise.

One useful framework breaks this decision into two stages:

  1. Connection trade-offs: range, power efficiency, and bandwidth
  2. Network-level trade-offs: cost, scalability, and integration complexity

By first identifying the type of connection your device requires, then refining based on deployment and business priorities, you can narrow your selection to the most suitable technology.

Connection Tradeoffs

The first major decision when selecting a wireless technology is finding the right balance between range, power efficiency, and data rate. These three factors often pull in different directions, and the best solution depends on your product’s intended use.

Short-range technologies such as Wi-Fi and Bluetooth Low Energy (BLE) are widely adopted for their high data throughput and compatibility with existing infrastructure, especially in consumer and industrial environments. These technologies are particularly effective for applications where the device operates within a defined, often indoor range, such as wearables, smart home devices, or factory floor sensors.

When longer range and low power consumption are more important than bandwidth, Low-Power Wide Area Network (LPWAN) technologies become a compelling alternative. Protocols like LoRaWAN and NB-IoT are optimized for low-energy communication across large areas and are often used in asset tracking, agricultural sensors, and utility meters. However, they typically trade off higher data rates and real-time communication capabilities to achieve this efficiency.

Traditional cellular networks like 4G LTE and 5G fill the need for high-speed, high-reliability communication across broad geographic areas. They are well-suited for mobile or mission-critical deployments that require robust Quality of Service (QoS) and low latency. The tradeoff, however, is significantly higher power consumption and recurring network costs, which makes them less suitable for battery-powered devices or cost-sensitive applications.

The diagram below illustrates these trade-offs visually, helping highlight how each category fits across the axes of power use, data rate, and coverage.

A three-circle Venn diagram comparing wireless connection types—Wi‑Fi/BLE/Zigbee, NB‑IoT/LTE‑M/Sigfox/LoRa, and 2G–5G cellular—by tradeoffs between low energy consumption, high data rate, and wide area coverage.

Network Tradeoffs

Once you’ve selected the appropriate connection category—whether short-range, LPWAN, or cellular—the next step is to refine that choice by considering the broader network-level trade-offs, particularly around cost, scalability, and security.

For example, cellular networks (like LTE-M and NB-IoT) generally provide the most reliable Quality of Service (QoS) and strong security guarantees, but they also bring higher hardware costs, subscription fees, and potentially more complex integration requirements. These solutions are most viable for large-scale, revenue-generating products where network resilience and global reach justify the investment.

In contrast, short-range solutions like Wi-Fi, BLE, and Zigbee offer low-cost implementations that can run on existing infrastructure. These are especially attractive for products with tight cost constraints or limited deployment range. However, they typically require some local infrastructure and are less scalable across distributed deployments without additional gateway architecture.

LPWAN options like LoRa and Sigfox often sit in the middle. These networks can offer strong coverage and battery life with moderate infrastructure costs, depending on whether the network is private (self-managed) or public (provider-managed). Public LPWANs reduce deployment effort but may introduce limitations in customization, QoS, and long-term control. Private LPWANs allow for more control and optimization but require greater up-front investment.

Evaluating the total cost of ownership (TCO) means looking beyond the module price. You also need to consider development effort, network setup and maintenance, and recurring data service charges. These factors can quickly shift the balance when comparing network options. What seems low-cost at the hardware level may become expensive over time, depending on your product’s update needs, data payloads, and deployment scale.

The following diagram illustrates how various wireless technologies map across trade-offs in cost, scalability, and security, offering another lens to guide network decision-making.

A three-circle Venn diagram illustrating tradeoffs between cellular, private, and public proprietary networks—such as Wi‑Fi, Zigbee, BLE, LoRa, and Sigfox—highlighting factors like security, scalability, and cost considerations.

Short Range Wireless

Short-range wireless is one of the most active and diverse areas in IoT connectivity, with many options for enabling private local networks and extremely low-power point-to-point communication. This section outlines the most widely used short-range technologies—BLE, Wi-Fi, and Wi-Fi HaLow—alongside alternatives that may be better suited for certain use cases.

While many of these technologies also support mesh networking, which enables devices to relay data across longer distances and build resilient topologies, this guide focuses on their core use in point-to-point or hub-and-spoke communication.

Bluetooth Low Energy (BLE)

BLE has become the cornerstone of modern short-range IoT communication. Its low-power profile and near-universal support on smartphones make it ideal for products that rely on mobile devices as both user interfaces and gateways to the cloud.

In practice, BLE strikes an ideal balance between energy efficiency and connectivity for applications like wearables, health monitors, asset trackers, and smart home products. One of BLE’s biggest advantages is its ability to communicate directly with mobile devices, eliminating the need for dedicated hubs or infrastructure.

At Punch Through, we’ve worked extensively with BLE in real-world applications and can confidently say it’s only becoming more ubiquitous in the connected product space.

Many modern BLE chipsets now offer Wi-Fi coexistence, allowing both protocols to run on the same radio. As a result, starting with a BLE + Wi-Fi combo has become a widely recommended baseline for connected product development. This combination enables low-power, device-to-device interaction via BLE, paired with high-throughput internet access via Wi-Fi without additional complexity.

Wi-Fi

Wi-Fi remains the most familiar and broadly deployed wireless standard, prized for its high bandwidth and direct internet readiness. Its global adoption and ease of integration make it an obvious fit for applications that require continuous connectivity, especially for devices that are stationary and powered.

In today’s ecosystem, selecting a Wi-Fi solution that does not also include BLE is becoming increasingly rare and in most cases, inadvisable. Combo chipsets offering both BLE and Wi-Fi simplify development, reduce infrastructure needs, and enhance user experience.

Wi-Fi is best suited for mains-powered or plug-in devices like home automation hubs, industrial sensors, and media streaming hardware. However, its relatively high power consumption makes it less ideal for battery-powered use cases requiring long lifespans.

Wi-Fi HaLow

Wi-Fi HaLow (IEEE 802.11ah) is an emerging option designed to address some of traditional Wi-Fi’s limitations for IoT. Operating in the sub-GHz spectrum, it offers longer range and better obstacle penetration, with lower power consumption than conventional Wi-Fi.

Unlike Zigbee and other constrained IoT protocols, Wi-Fi HaLow is IP-native, meaning it integrates directly with existing internet infrastructure, no translation layers required. That makes it especially well-suited for industrial, agricultural, and smart city applications where long-range, energy-efficient IP connectivity is critical.

Although adoption is still growing, HaLow’s interoperability and balance of coverage and efficiency make it a strong candidate for future-proof IoT deployments.

Short Range Alternative

For applications requiring low power and moderate range, Low-Rate Wireless Personal Area Networks (LR-WPANs) and some other short range protocols provide viable alternatives to BLE and Wi-Fi in specific cases. These networks are particularly common in smart home and building automation, point of sale systems or very specific consumer products, where devices need to communicate over short distances with minimal energy consumption.

  • Zigbee is the most widely adopted LR-WPAN protocol, known for its mesh networking capabilities and broad support among silicon vendors. Zigbee is a strong choice for smart lighting, home security, and industrial monitoring applications.
  • Z-Wave is another prominent smart home protocol, operating in the sub-GHz range for improved range and lower interference compared to Zigbee. It is widely used in home automation but has a smaller ecosystem.
  • ANT is a specialized ultra-low-power protocol designed for sports and fitness applications, commonly found in heart rate monitors and cycling sensors.
  • EnOcean is a unique energy-harvesting wireless standard that eliminates the need for batteries, making it ideal for low-power, maintenance-free IoT devices.
  • RFID provides short-range, low-cost communication primarily for asset tracking and access control applications.

Low-Power Wide Area Networks (LPWAN)

Low-power wide area networks (LPWANs) are often the most effective solution for applications that require both long-range connectivity and low power consumption. These technologies are designed to support large-scale, low-bandwidth IoT deployments, which are common in sectors like utilities, agriculture, logistics, and infrastructure monitoring.

LPWANs can be broadly divided into two main categories, each with distinct strengths and trade-offs:

  • Cellular-based LPWANs (e.g., NB-IoT, LTE-M) leverage existing mobile networks to deliver wide coverage and high scalability. These are ideal for geographically dispersed products and applications where quality of service and reliability are critical. However, they typically come with higher module costs and ongoing network subscription fees, which can add up over time, especially in high-volume or price-sensitive applications.
  • Non-cellular LPWANs (e.g., LoRaWAN, Sigfox) offer a lower-cost alternative with minimal infrastructure and subscription requirements. These solutions are often a better fit for region-specific, battery-powered, or lower-volume deployments. LoRaWAN, in particular, supports both public and private network architectures, offering additional flexibility depending on deployment needs. Sigfox, while simple to implement, is dependent on existing public infrastructure and has seen more limited global adoption.

Ultimately, the choice between cellular and non-cellular LPWAN technologies comes down to your product’s coverage needs, data throughput, cost sensitivity, and infrastructure control requirements.

Traditional Cellular

For products that require wide-scale geographic coverage and high data throughput, cellular broadband (e.g., 4G and 5G) is often the only viable option. These technologies are well-suited for mobile or field-deployed devices such as fleet tracking systems, logistics platforms, and remote sensing applications.

However, cellular broadband comes with notable trade-offs:

  • Higher network costs per device due to subscription fees
  • More expensive modules, making them cost-prohibitive for low-volume or price-sensitive products
  • Increased power consumption, making them unsuitable for battery-powered devices requiring long lifespans

Given these constraints, cellular broadband is typically best reserved for use cases where no other technology can meet the required combination of range, reliability, and bandwidth. While it delivers robust coverage and performance, those benefits often come at a premium.

Connectivity Selection Recap

For most connected product applications, Bluetooth Low Energy (BLE) and Wi-Fi remain the most reliable and scalable defaults. They both support local and cloud communication while limiting power use and development complexity.

When your product demands a longer range or ultra-low power, LPWAN and LR-WPAN technologies offer viable alternatives. However, each brings trade-offs in network architecture, cost, and deployment control. Broadband cellular, while powerful, should be treated as a niche solution best reserved for data-heavy or mission-critical deployments where nothing else fits.

Emerging options like Wi-Fi HaLow are worth keeping an eye on, particularly if your product roadmap includes more distributed, infrastructure-light use cases. But adoption should be weighed carefully against maturity and integration overhead.

No matter the technology, the most resilient connectivity strategies are grounded in real-world constraints. Prioritize documentation quality, protocol compatibility, and deployment readiness over theoretical performance. When in doubt, lean on comparative testing, protocol standards, and firsthand development experience to make the call.


Considering the Embedded Operating System

Your choice of embedded OS will shape nearly every aspect of your connected product from memory and power constraints to security architecture and update mechanisms. It influences not just how your team builds, but what you can build, how fast you can iterate, and how well your product holds up post-launch.

For most products, the decision comes down to two paths: 

  1. A lightweight RTOS that prioritizes simplicity and efficiency. 
  2. Embedded Linux, which offers flexibility and power at the cost of increased complexity.

Neither is inherently better. But picking the wrong one can slow your roadmap, balloon your BOM, or create hidden security and maintenance debt. This section explores how to make the right call based on your product’s goals.

Comparing RTOS and Linux

RTOS

RTOS-based systems offer lower power consumption and minimal resource usage, making them ideal for constrained environments. They provide real-time performance capabilities with a smaller memory footprint, allowing for cost-effective component selection. The software structure is simpler, which reduces security overhead and lowers the overall bill of materials (BOM). 

Direct low-level hardware access is another advantage, enabling efficient interaction with peripherals. At PT we typically standardize development around FreeRTOS, but there are other embedded RTOS solutions available, most notably Zephyr.

However, RTOS solutions come with limitations. They often have restricted library and connectivity stack support, making it challenging to implement advanced features. Available libraries and SDKs may be poorly documented or have feature constraints, requiring additional effort for development.

Linux

Embedded Linux simplifies network connectivity implementation and offers extensive support for device drivers, including cameras and displays critical for interactive UIs. It provides a broader selection of programming languages, access to a rich ecosystem of libraries and connectivity stacks, and readily handles the processing power needed for demanding applications like AI. 

Additionally, non-recurring engineering (NRE) costs for software development can often be lower thanks to this extensive ecosystem.

Despite these advantages, Linux introduces significant complexity. Its software stack, including the bootloader, kernel, and filesystem, is considerably more intricate than RTOS-based designs. Configuring Linux can be difficult, and securing a Linux-based system requires greater effort compared to an RTOS. 

Higher resource requirements typically result in increased BOM costs, and importantly, designing the more complex hardware often needed to run Linux can lead to higher hardware NRE costs compared to simpler RTOS systems.

 Moreover, connection security complexities, particularly around device key protection, add another layer of challenges. Over-the-air (OTA) updates also demand robust security mechanisms and become increasingly difficult as security needs grow.

Which OS Is Right for Your Product?

An RTOS is the safest default for most connected products, offering strong capabilities without overcomplicating development. Solutions like Zephyr strike a practical balance between resource efficiency, predictable performance, and maintainable software structure.

In our experience, teams building RTOS-based systems gain more long-term leverage by mastering lightweight embedded stacks than by managing the overhead of embedded Linux. And increasingly, we see the market leaning toward low-power, cost-sensitive designs that simply don’t justify the complexity of a Linux-based system.

That said, Linux can be the right fit, particularly for applications that require advanced networking, rich UI support, or fast prototyping with off-the-shelf modules. If your product’s business model and power constraints allow for it, Linux may offer faster time to market and broader ecosystem support.

Ultimately, this decision hinges on your product’s requirements. The wrong call can slow development, introduce unnecessary security and integration work, or inflate your BOM. The right one creates a streamlined path to launch with fewer surprises down the road.


Considering IoT Application Protocols and Cloud Compatibility

Communication protocols are often overlooked in hardware selection, but mismatches here are one of the most common (and costly) causes of integration delays.

This risk is especially high with RTOS-based systems, where protocol support is less abstracted and often tightly coupled to the hardware platform. If your device’s messaging stack doesn’t align with your cloud provider—or if key libraries aren’t available for your chosen platform—you may face custom workarounds, added security concerns, or late-stage architectural changes.

Most cloud platforms are optimized around specific protocols like MQTT, HTTP, or CoAP. Choosing a hardware platform that supports those protocols natively can reduce development time, simplify OTA updates, and prevent long-term maintenance friction. It’s not just about connectivity, it’s about aligning your communication stack with your business model and your go-to-market timeline.

Application Protocols

IoT application protocols can be broadly categorized into two groups:

  1. Messaging Protocols: Handle communication between devices and cloud platforms.
  2. Network Management Protocols: Facilitate device management tasks like setup, firmware updates, and monitoring.

Messaging Protocols

Messaging protocols define how IoT devices exchange data with cloud platforms and other devices. Key factors to consider when choosing a messaging protocol include:

  • Latency: Response time of message delivery.
  • Throughput: Volume of data that can be transmitted.
  • Energy Consumption: Impact on battery life for low-power devices.
  • Communication Model: Whether the protocol supports publish-subscribe or request-response paradigms.

According to A Survey on Communication Protocols and Performance Evaluations for Internet of Things, messaging protocols for IoT devices differ widely in latency, throughput, power consumption, and communication model. These differences matter especially in constrained environments where efficiency and reliability are critical.

When selecting hardware, it’s essential to ensure that the device supports your chosen protocol, both in terms of resource requirements and available software stack compatibility.

The most common IoT messaging protocols include:

  • MQTT: Optimized for lightweight communication, particularly in scenarios where devices need to send messages to multiple subscribers.
  • CoAP:  Provides high efficiency with minimal overhead by sacrificing handshaking mechanisms, making it ideal for constrained networks.
  • HTTP: Commonly used but less efficient for IoT applications due to its stateless nature.
  • AMQP: Suited for enterprise messaging applications requiring robust messaging queues.
  • DDS: Supports real-time data exchange, often used in mission-critical systems.
  • WebSockets: Ideal for high data rate applications requiring persistent connections.

Specialized Network Management Protocols

Specialized network management protocols enable advanced remote configuration, monitoring, and updating of IoT devices. While messaging-based device management (typically MQTT-based) is common, specialized protocols may be required for specific device management needs or to accommodate low-power constraints. 

These protocols play a crucial role in device lifecycle management but may have limited support in embedded systems. Therefore, it is essential to verify that a hardware platform supports the required network management protocol. The following includes some examples of network management protocols, some being more agnostic to the underlying connection technologies (i.e. 6LoWPAN-SNMP) than others (i.e. LwM2M).

  • LwM2M: Lightweight protocol designed for remote device management and firmware updates.
  • CoMI: Compact protocol for device configuration using CoAP.
  • 6LoWPAN-SNMP: Tailored specifically for 6LoWPAN IPv6-based low-power networks.
  • NETCONF Light: A streamlined version of NETCONF for constrained devices.

Cloud Platform Compatibility

The availability and support for IoT application protocols vary across cloud platforms. Most major platforms support MQTT and HTTP, while fewer offer native support for specialized protocols like CoAP, DDS, and the more obscure network management protocols. 

Furthermore, some platforms provide fully integrated IoT solutions, from embedded stacks to cloud services, while others offer more limited support, possibly necessitating third-party integration efforts to support the right feature. A comprehensive comparison of various IoT cloud platforms, as detailed in A Review of IoT Network Management: Current Status and Perspectives, highlights how not all cloud platforms offer the same services and capability.

By carefully evaluating cloud platform compatibility during hardware selection, you can avoid integration issues, avoid extra development effort and realize a simpler network management solution.

Key Considerations for Cloud-Compatible Hardware Design

To avoid integration issues, rework, and unnecessary complexity, your hardware selection process should explicitly account for cloud compatibility. Here’s what to evaluate:

  • Protocol Support: Does the hardware natively support the protocols your application needs (e.g., MQTT, HTTP, CoAP)? This is especially important for RTOS-based systems, where protocol libraries are often limited or tightly coupled to the hardware SDK.
  • Cloud Platform Fit: Will your chosen cloud provider support those protocols without custom translation layers or third-party tooling? Not all platforms support CoAP or LwM2M out of the box.
  • Developer Ecosystem & Integration Maturity: Does the cloud provider offer native SDKs, device management services, or streamlined onboarding for your hardware stack? Strong vertical integration can reduce development time and maintenance effort.

Default recommendation: Design with support for MQTT and HTTP unless your power constraints demand alternatives like CoAP or LwM2M.

Before finalizing hardware, validate that:

  • Your selected platform includes embedded support for your required protocols
  • Your cloud provider offers native support for those same protocols

Without alignment between your hardware and cloud platform, you risk introducing avoidable complexity like extra middleware, security vulnerabilities, and lost development time.


Comparing IoT Chip Vendors

The chip vendor you choose influences more than just component selection. It affects your team’s ability to move fast, scale confidently, and avoid integration challenges. Their ecosystem, documentation, and support model directly shape your development experience and long-term success.

In our experience, the most successful teams evaluate chip vendors across four key dimensions, not just technical specs:

  • Connectivity Capability: Does the vendor support a range of wireless protocols (e.g., Wi-Fi, BLE, LoRa, cellular) that allow your product to scale or pivot without changing hardware?
  • Documentation Quality: Can your team move quickly and independently? Poor documentation slows development, increases bugs, and burdens onboarding.
  • Software Ecosystem Support: Does the vendor provide well-maintained SDKs, RTOS integrations, and middleware? This reduces the need for custom code and simplifies security and connectivity.
  • Part and Dev Kit Availability: Are modules and kits readily available for prototyping and production? Limited access can introduce delays and supply chain risk.

Vendors that deliver consistently in these areas reduce friction at every stage of development, from proof of concept to post-launch support. The next section provides a high-level comparison of leading vendors based on these factors.

Comparing Vendors

Vendor selection plays a key role in the long-term success of an IoT product. In addition to evaluating core technical capabilities, it’s essential to assess each vendor’s documentation quality, software ecosystem, and part availability, all of which can significantly impact development speed, risk, and scalability.

That said, no single vendor is the best fit for every application. Some specialize in low-power wireless, others in embedded Linux performance, and others still in cellular IoT connectivity. The best fit depends on your specific design and business priorities.

Below is a quick summary of leading IoT silicon vendors, followed by a table visualizing their strengths across the four key criteria.

Vendor Snapshot

  • Qualcomm: Best for cellular IoT (LTE-M, NB-IoT, 5G), with strong Wi-Fi and Bluetooth support. Excellent documentation, but part availability can be limited for small runs.
  • Nordic Semiconductor: Ideal for low-power wireless solutions (BLE, Thread, Zigbee). Well-documented, actively supported, and widely available.
  • Semtech: Market leader in LoRaWAN for long-range, low-power IoT. Solid SDK support, but availability has been affected post-Sierra Wireless acquisition.
  • Espressif: Great for budget-friendly Wi-Fi/Bluetooth applications. Excellent documentation and tooling, though some workarounds may be required with certain stacks.
  • Quectel: Focused on cellular IoT with broad support for LTE-M, NB-IoT, and 5G. Hardware is easy to source, but software support and documentation are lacking.
  • Telit: Offers strong cellular options for industrial applications, but dev kits are limited and much of the documentation is gated.
  • Silicon Labs: Versatile vendor with BLE, Zigbee, Thread, and Wi-Fi. Known for excellent documentation and broad toolchain support.
  • NXP: Strong in secure edge computing, with support for multiple protocols (Wi-Fi, BLE, Zigbee, Thread). Good documentation and hardware availability.

This table offers a side-by-side view of vendor strengths, helping teams identify which options align best with their technical and operational priorities:

ManufacturerIoT CapabilityDocumentationSoftware SupportAvailability
Qualcomm🟢🟢🟢🔴
Nordic🟢🟢🟢🟢
Semtech🟡🟡🟢🔴
Espressif🟡🟢🟡🟢
Quectel🟢🔴🔴🟢
Telit🟢🟡🟡🟡
Silicon Labs🟡🟢🟢🟢
NXP🟡🟢🟢🟢

Why it Matters

The right chip vendor affects how quickly your team can build, how easily you can integrate new features, and how much risk you carry through development and production. Poor documentation and immature SDKs slow teams down. Limited part availability introduces delays or forces late-stage design changes. Weak integration support leads to brittle implementations that break under real-world conditions.

A strong vendor relationship built on solid tooling, clear documentation, and responsive support reduces friction across the board. It enables faster prototyping, more stable production code, and smoother updates over time. In a fast-moving IoT space, that level of support is what keeps your product competitive.


The Verdict: Designing a Scalable Hardware Foundation

Hardware decisions shape the rest of your product: how it performs, how long it takes to build, and how well it can evolve. From power strategy and connectivity to security and cloud integration, everything depends on getting this foundation right.

Across dozens of connected product projects, we’ve seen a few patterns hold up:

  • Connectivity: BLE + Wi-Fi remains the most practical and flexible default for short-range IoT applications.
  • Operating System: An RTOS (like FreeRTOS or Zephyr) should be the default unless your application justifies Linux’s added complexity.
  • Cloud Protocols: MQTT and HTTP provide broad compatibility and efficient communication for most use cases.
  • Vendor Fit: Prioritize vendors with strong documentation, developer support, and reliable part availability as they directly affect your development velocity and risk exposure.

A good default platform minimizes complexity and leaves room to adapt when your product demands it. These aren’t just component choices, they’re architectural leverage points that define how you build and scale.

Recommended starting points:

  • Espressif ESP32-C6: Budget-friendly BLE + Wi-Fi module with strong community support
  • Silicon Labs SiWx917M: Flexible dual-protocol platform with excellent tooling
  • NXP RW612: Versatile module with robust security features
  • Nordic nRF5340 + nRF7002: Best-in-class low-power BLE with separate Wi-Fi integration

When to Bring in a Partner

These early choices often look straightforward until real-world complexity sets in. If your team is facing unclear trade-offs between platforms or protocols, gaps in embedded expertise that slow progress, or pressure to move fast without adding risk, it might be the right time to bring in a partner. We’ve helped teams simplify complex trade-offs, avoid costly missteps, and move forward with confidence, and we can help you do the same. Reach out if you’d like to learn more.


Final Thoughts: Building With Confidence

Hardware isn’t just a technical layer, it’s the backbone of your connected product. The platform decisions you make now will ripple through your system architecture, development velocity, and long-term maintainability.

If there’s one core takeaway, it’s to optimize for clarity, not theoretical maximums. The best platforms balance your product’s actual constraints (power, cost, time to market) with a solid ecosystem and scalable architecture. That’s what sets you up to deliver reliably and iterate with confidence.

Share:

Garin Marlow
Garin Marlow
Garin develops firmware for connected devices, especially in the medical space. He enjoys solving complex technical challenges and building systems that make an impact. In his free time, he watches pro disc golf, plays volleyball, and builds fun side projects like custom Mario Kart courses.

Subscribe to stay up-to-date with our latest articles and resources.