Apple Watch or Wear OS: which is the best platform to build on?
It depends on your target audience and your existing mobile app. Apple Watch has a strong health framework (HealthKit), a more loyal user base and better-integrated complications, but only works with iPhones. Wear OS (Samsung, Google, Mobvoi) has a growing share, covers Android users and offers more hardware choice. For consumer and lifestyle apps, we often recommend both. For a closed population (clinical research, a sports team, an internal workforce), a single platform is more efficient.
Which programming languages and frameworks do you use?
For watchOS, we build natively in Swift with SwiftUI and WatchKit. For Wear OS, we write in Kotlin with Jetpack Compose for Wear. Garmin Connect IQ requires Monkey C, a proprietary language. For AR wearables such as HoloLens or Magic Leap, we work in Unity, sometimes WebXR for lighter experiences. Cross-platform Flutter for wearables has limitations around complications and background sensors, so for health or industrial apps we tend to choose native.
We want to bring a medical wearable product to market: what do we need to arrange?
Software on a medical device falls under CE-MDR (Medical Device Regulation), or CE-IVDR for in-vitro diagnostics. The software class determines how demanding the process is: a wellness tracker is a different matter from a sensor that supports treatment decisions. We work with your regulatory team or an external consultant on the technical documentation (IEC 62304), risk management (ISO 14971) and the clinical evaluation. NEN 7510 and the GDPR go hand in hand if you process patient data. See also our
healthcare software practice.
How do you make sure the battery doesn't run flat within half a day?
The battery budget is a design decision, not an optimisation after the fact. We measure early on real hardware to see which functions cause the biggest drain: continuous BLE connection, GPS, screen wake, sensor polling. The typical techniques: syncing intermittently instead of streaming continuously, low-power modes with cached state, sensor fusion so we need to sample less often, and shifting work to the phone where possible. For clinical devices, we also build hardware-side firmware recommendations into the design.
How reliable is the BLE connection between wearable and phone?
BLE is robust, but not always predictable. Disconnects happen in lifts, in MRI environments, when hands are full or batteries are weak. We design every wearable app on the assumption that the connection can drop at any moment: local buffering on the device, retry logic with backoff, conflict resolution when both sides have been updated, and clear UI feedback on the connection state. For industrial and healthcare contexts, we also run explicit reliability tests on site.
Can you also handle AR glasses such as HoloLens or Magic Leap?
Yes. For AR wearables, we work in Unity with Microsoft's MRTK toolkit or the respective SDKs for Magic Leap and Apple Vision Pro. Use cases we come across regularly: AR work instructions for technicians or operators, remote-expert sessions where a specialist watches through the glasses, and visualisation of 3D data from an ERP or digital twin. We treat Vuzix and RealWear as Android devices with a UX adapted for voice control, which is a different class from immersive AR.
How do you handle health data and privacy?
Health data is special category personal data under the GDPR and requires strict data processing agreements. We apply privacy by design: minimising what the wearable stores and sends, end-to-end encryption where possible, explicit opt-in flows and logging that meets NEN 7510. For healthcare organisations, we sign a data processing agreement and carry out a DPIA at the start. We do not store patient data in our own environments unless that is explicitly part of the scope.
What determines the cost and lead time?
The biggest factors: the number of wearable platforms (watchOS-only is cheaper than watchOS + Wear OS + Garmin), whether a new companion app has to be built or an existing app extended, how many sensor types and cloud integrations are added, and whether certification (CE-MDR) is part of the scope. A well-defined smartwatch app with a companion is a project of a few sprints; a full medical wearable stack with clinical integration is a project of multiple sprints. We're happy to work out an initial indication during our first conversation.
An existing wearable from the market, or design your own hardware?
In most cases, we advise first checking whether an existing wearable (Apple Watch, Garmin, a certified medical sensor from Withings or Cardiomo, an industrial headset from RealWear) covers your use case. Hardware development is a discipline of its own, with its own risks. Only if existing products truly fall short do we look at custom hardware together with a hardware partner, while we focus on the firmware interface, the companion app and the cloud platform.