Inside the Engineering Stack - Why Building Smart, Secure Wearables Remains a Hard Engineering Problem

Home » Insights » Inside the Engineering Stack – Why Building Smart, Secure Wearables Remains a Hard Engineering Problem

The wearables market is moving from land-grabbing to a steadier, more strategic pace. Smartwatch volumes normalize as pricing pressure persists and replacement cycles lengthen. Lighter entrants (smart rings and glasses) slip into open lanes with focused use cases.

For technology leaders, this is less of a gadget story than an engineering economy - finding the right wearable embedded software development approach. Generative AI and richer connectivity raise expectations at the edge, but they only scale with disciplined embedded software.

In wearables, the race is not just about who ships the most features the fastest, but who can turn constraints into advantages – battery budgets, RF topologies, and model footprints. These engineering decisions, together with choosing the right wearable software development partner, will determine who breaks away in the next stage.

This article explains the technical realities of building AI-ready wearable devices by orchestrating three constraints – power, latency, and trust – through edge/on-device AI, disciplined power budgeting, and secure-by-design updates. It discusses why many wearables PoCs stall and offers in-depth perspectives into real challenges and solutions from rinf.tech’s development experience with wearables. Finally, it scans predictions and emerging trends in the wearables market.

What's the state of the wearables market in 2025?

After reaching 534.6 million units in 2024, the global wearables market is expected to experience a more moderate growth trajectory of 4.1% in 2025. Yet it has entered another transformative year, driven by technological advancements and evolving consumer needs.

After a challenging 2024 when smartwatches recorded a 4.5% decline primarily due to an influx of low-cost options, 2025 debuted with elevated inventory levels and debates regarding price strategies and overall sustainability. In the US, the market contracted due to the large number of devices purchased during the pandemic and the resulting customer continuity in using them. An upgrade of older devices is expected between 2026-2027, which is said to influence a 0.9% growth in overall smartwatch shipments. On a global scale, smartwatch shipments are expected to grow 2.5% in 2025, according to Wearables Devices Market Insights by IDC.

The same research cites an impressive 2024 for hearables, recording a growth of 8.9%, and set to remain the largest category within the market, continuing to grow in the next few years. Emerging markets create a favorable context, with regular refresh cycles, reductions in average selling prices, and ongoing innovations such as open-ear designs.

New entrants are opening doors to new possibilities in the market, as vendors consider them alternative tracking devices for various uses. Some experts predict that smart rings and glasses without displays will be the most interesting categories in the coming quarters.

As more first-time buyers explore wearable technology, the industry is moving beyond mere data tracking to redefine personal tech experiences, offering health and wellness coaching through intelligent algorithms and real-time insights. Generative AI, enhanced connectivity, and growing consumer interest in areas like continuous health and fitness tracking are driving new endeavors in this dynamic market. This rapid evolution signals an urgent demand for robust embedded software for wearables to support advanced functionalities and diverse applications of next-generation devices.

Core Engineering Constraints in Wearable Product Development

Building next-gen wearables is ultimately a game of three unforgiving budgets: energy, time, and trust. This is especially true in wearable firmware development, where power, latency, and update hygiene define real-world success.

1. Power Budget

Embedded systems must efficiently process and transmit complex data in real time while maintaining low power consumption. Real-time machine learning models, crucial for advanced functionalities, must be lightweight to run on low-power devices, often necessitating techniques such as model compression and pruning. The industry’s pursuit of ultra-slim designs introduces high-density battery thermal runaway risks, which can lead to increased design iterations, delayed launch timelines, and higher supply-chain costs.

Overall, short battery life remains a general concern in wearable product development. This makes low-power wearable design and power-budget optimization a first-class requirement for any project.

Bogdan Popescu, Technical Director at rinf.tech within the R&D Embedded Systems division points out that “In wearable devices, it is essential that the operating system is configured to activate only those kernel components that are absolutely necessary, such as memory management, the process scheduler, and the I/O system.” According to Bogdan, “This ensures minimal resource consumption and maximizes battery life, while still allowing applications to run smoothly enough to be usable. For example, if the system consumes too many resources, an ECG monitoring app would only be able to take readings at a very low frequency, resulting in significant inaccuracies. In addition, when reading values from various sensors, there are two possible interface types: SPI and I2C. In some of the projects we developed, tests were conducted using both types, and the accuracy of the readings was found to be identical. Based on this, SPI was selected because it allows the sensor to consume less power than with I2C – an important factor for extending battery life.”

2. Latency

Real-time data collection and analysis are crucial for providing immediate feedback, coaching, education, and remote support. Edge AI reduces dependence on the cloud for the fast path, improving responsiveness and resilience. Advanced connectivity solutions, such as 5G Stand-Alone modules, can support low-latency streaming, but end-to-end results vary by network and deployment.

3. Security Updates and Data Privacy

Wearable devices collect highly sensitive physiological and behavioral data, which raises significant concerns regarding privacy and security. Robust encryption, anonymization, and strict compliance with regulations such as HIPAA in the US and GDPR in the EU are essential.

From day one, treat secure OTA updates for wearables and a maintained SBOM for wearables as non-negotiable parts of the architecture.

4. Development of On-Device Applications for Small Displays

Finally, Andrei Hutuca, R&D Embedded Systems Technical Delivery Manager at rinf.tech, points to another challenge associated with wearables development projects – the development of on-device applications for small displays: “To address the limitations of small display sizes, two categories of applications must be developed for these devices: apps that run directly on the device and apps that run on a smartphone paired with the device. On-device applications must be adapted to the small screen size (approximately 3×3 cm) and the resolution supported by the device to ensure optimal usability. This requires a process of experimentation and adaptation in the way applications are designed for the device.”

According to Andrei, “For smartphone-based applications, the main challenge lies in how the phone connects to the device and how smoothly the app can extract data from it without causing any noticeable interruptions for the user. Another technical difficulty, influenced by connection quality and sensor data accuracy, is the need for apps to include mechanisms that detect and handle situations where received data is erroneous, or to filter out values considered outside the normal range for a specific user. These mechanisms must be personalized for each user and cannot rely on standard thresholds – as every individual is different.“

Why Wearable PoCs and Advancements Stall

Last Mile Engineering

The urgent need for advanced R&D in embedded systems, particularly as devices become more intelligent and interconnected, often encounters critical bottlenecks, such as the ones described earlier. These challenges frequently lead to promising Proof of Concepts (PoCs) stalling before market readiness. This is where a seasoned wearable software development partner can de-risk the handoff to production.

As the wearables market remains massive and is expected to grow steadily through 2027 across segments, only wristbands are projected to slow down in the coming years. Thus, a core tension continues: product leaders want more on-device intelligence but must sustain multi-day battery life. As a result, PoCs or product advancements typically stall at the last-mile engineering step: prototype-to-production wearables that are certifiable, maintainable, and secure.

Bridging this gap involves rigorous testing, adherence to complex regulatory compliance, advanced optimization for power and performance, and seamless integration across diverse hardware and software components. Projects can falter at this juncture due to a lack of specialized expertise in areas such as ultra-low power design, robust security implementation, or navigating the intricate landscape of global certifications. This final, often most challenging phase demands precision and deep technical acumen to mitigate risks, accelerate time-to-market, and ensure the successful productization of complex wearable innovations.

Challenges Leading to PoCs Stalling and Hindering Product Advancement

Future Outlook: What’s Ahead

Milestone 1: Bluetooth 6.1 and 6.x (2025 – 2026)

Impact: Better ranging (Channel sounding), privacy/power refinements, and a bi-annual release schedule that should yield incremental improvements through 2026.

Bluetooth 6.0 (late Aug/Sep 2024) introduced Channel Sounding (cm-level ranging), and new filtering among other features. Bluetooth 6.1 landed in May 2025 with privacy/power updates and a bi-annual release schedule going forward. The market expects to see further 6.x increments into 2026 that can further improve location and efficiency.

Milestone 2: EU Cyber-Resilience Act Enforcement (Expected: 2027)

Impact: Mandatory secure update capability and SBOM/vulnerability handling. Practical steps toward compliance for wearables.

By mandating that manufacturers and retailers maintain cybersecurity throughout the lifecycle of their products, the Cyber Resilience Act raises the bar for cybersecurity standards in the EU for goods with digital components, from smartwatches to baby monitors. The act entered into force on December 10, 2024, with main obligations taking effect from December 11, 2027.

The Cyber Resilience Act mandates new cybersecurity standards for all digital products, except for some specific exemptions. This regulation applies to manufacturers and retailers, covering every step of a product’s life cycle, from its initial design to its ongoing maintenance. It impacts all devices that can connect to a network or another device. Products already covered by other regulations, such as those in the medical, aviation, and automotive industries, are excluded, as is some open-source software.

Milestone 3: More Sub-Segments Identified as Growing Product Category (2025 – 2030)

Impact: Continued Product Diversification

Advancements in technology are making subcategory products, such as smart glasses, a practical and popular option. This trend is expected to continue as technology improves, more consumers become aware of the products, and new companies enter the market.

In 2025, IDC forecasts 18.7 million units in 2029, compared to 2.7 million units in the most recent full year of data, 2024. IDC evaluates that the current growth trend is marked by rapidly improving technology, enabling the release of glasses packed with features while remaining relatively lightweight. The recent results of this segment, according to IDC, may signal that consumers are now viewing smart glasses as a legitimate technology product with significant real-world use cases. Even if eyewear still has a long way to go, an increased appetite for hands-free and AI-assisted interfaces will contribute to raised expectations across product categories. Furthermore, the rise of rings and lightweight eyewear also increases demand for smart ring software and smart glasses software tuned to tiny displays and batteries.

Conclusions

Complex challenges, including stringent power constraints, the imperative for advanced security, and the intricate integration of sophisticated AI at the edge characterize the development of next-generation wearable embedded software. The persistent “last-mile engineering” gap, where promising proof-of-concepts often falter before achieving market readiness, represents a significant hurdle for tech leaders.

Advancements in Bluetooth connectivity are expected to enable sophisticated functionalities, and along with AI-on-chip wearables, will significantly broaden the addressable market to include more use cases and users. Embedded software, crucial for managing sensor inputs, processing data efficiently, and ensuring secure encryption, will continue to be a foundational element for advanced capabilities.

At rinf.tech, we have a long history of working with companies whose goal is to transform bold ideas into certified, successful wearable products. Whether you’re pushing the limits of edge AI, navigating power constraints, or preparing for upcoming regulations, our experts within the specialized R&D Embedded division at rinf.tech can support you in building smarter, safer devices, from PoC to mass production. Our ability to handle complex software engineering projects in wearables and provide strategic guidance is validated by long-lasting partnerships with some of the top market leaders.

To explore how we can accelerate your wearable product development, enhance security, and navigate the complexities of the future market, get in touch and let’s discuss your needs and how we can support.

Looking for a technology partner?

Let’s talk.

Inside rinf.tech’s 8-Year Partnership with Intel on Hardware-Software Integration and Open Source – Powering Cloud, IoT and AI

This article examines how rinf.tech’s 8-year partnership with Intel has advanced open-source innovation by addressing real-world challenges in cloud, IoT, and AI ecosystems. It highlights joint contributions to improve workload performance, enhance IoT sensor integrations, and optimize computer vision using OpenCV and OpenVINO. Read more into how these contributions deliver measurable value to the global developer community and accelerate enterprise adoption of scalable, reliable technologies.

September 11, 2025

embedded