Quantcast
Jump to content


Recommended Posts

Posted

2026-04-10-01-MainBanner.jpg

Health and fitness are the most popular features for Galaxy Watches running Wear OS powered by Samsung. Implementing these features requires a continuous data stream to work effectively and seamlessly.

One of the most common challenges third-party developers face is keeping a sensor like the heart rate monitor active, even when the watch screen is off.

By default, Wear OS efficiently optimizes power consumption to extend usage time. As part of this optimization, sensor data collection may stop when the screen is off. This presents a challenge for applications that require continuous monitoring, such as health trackers, workout assistants, or medical-grade wearables.

What Happens When the Screen Turns Off

When a Galaxy Watch screen turns off, the system enters a low-power state to preserve battery. During this time:

  • The CPU may slow down or suspend execution of background threads.
  • Registered sensor listeners can stop receiving updates.
  • Any active work in your Activity pauses.

To keep sensor data flowing, your application needs two things:

  • A foreground service to keep your code running in the background.
  • A wake lock to prevent the CPU from going into a deep sleep state.

This tutorial shows how to create a simple Galaxy Watch application that continuously collects heart rate data, even when the screen is off, using a foreground service and a wake lock and all the code examples are provided in a downloadable sample application.

Let's Start Coding

Step 1: Create a New Wear OS Project

Open Android Studio and create new project from scratch:

  1. Go to File > New > New Project > Wear OS Tab > Empty Wear App.
  2. Fill in the project details in the New Project window.

2026-04-10-01-01.jpg

Figure 1: Create a new wearable application in Android Studio

  1. Click Finish and wait for Gradle sync to complete.

Step 2: Configure Permissions in AndroidManifest.xml

In the AndroidManifest.xml file, add the following permissions to access the heart rate sensor, foreground service, and wake lock.

<uses-permission android:name="android.permission.BODY_SENSORS" />
<uses-permission android:name="android.permission.FOREGROUND_SERVICE" />
<uses-permission android:name="android.permission.FOREGROUND_SERVICE_HEALTH" />
<uses-permission android:name="android.permission.WAKE_LOCK" />
<uses-permission android:name=" android.permission.BODY_SENSORS_BACKGROUND " />
<uses-feature android:name="android.hardware.type.watch" />

Register the SensorService in the manifest:

<service
            android:name=".SensorService"
            android:enabled="true"
            android:exported="false"
            android:foregroundServiceType="health" />

Step 3: Design Your Watch UI Layout

The watch UI can be designed entirely according to your preference. In this content, only two buttons have been used to start and stop the service and a TextView to show the result to keep it simple. Wear OS screens are small, so keeping the layout simple is the best practice.

To implement the UI, edit app/res/layout/activity_main.xml.

The following code implements a sample UI:

<?xml version="1.0" encoding="utf-8"?>
<androidx.constraintlayout.widget.ConstraintLayout xmlns:android="http://schemas.android.com/apk/res/android"
    xmlns:app="http://schemas.android.com/apk/res-auto"
    xmlns:tools="http://schemas.android.com/tools"
    android:layout_width="match_parent"
    android:layout_height="match_parent"
    android:padding="16dp"
    tools:context=".MainActivity"
    tools:deviceIds="wear">

    <LinearLayout
        android:layout_width="match_parent"
        android:layout_height="match_parent"
        android:gravity="center"
        android:orientation="vertical">

        <TextView
            android:id="@+id/heart_rate_text"
            android:layout_width="wrap_content"
            android:layout_height="wrap_content"
            android:text="--"
            android:textColor="#90EE90"
            android:textSize="24sp"
            android:layout_marginBottom="16dp"/>

        <Button
            android:id="@+id/start_button"
            android:layout_width="wrap_content"
            android:layout_height="wrap_content"
            android:text="@string/start_sensors" />

        <Button
            android:id="@+id/stop_button"
            android:layout_width="wrap_content"
            android:layout_height="wrap_content"
            android:layout_marginTop="8dp"
            android:text="@string/stop_sensors" />

    </LinearLayout>

</androidx.constraintlayout.widget.ConstraintLayout>

Step 4: Edit MainActivity.java

Inside MainActivity.java, sensor permissions must be requested at runtime.


onCreate() :
You need to initialize all the UI components inside the onCreate() method. This example uses two Button instances, for starting and stopping the service, and one TextView , for showing the result. Before staring the service, you have to check all the runtime permissions.

heartRateTextView = findViewById(R.id.heart_rate_text); //initialize globally to use it outside of the method
        Button startButton = findViewById(R.id.start_button);
        Button stopButton = findViewById(R.id.stop_button);

        if (startButton != null) {
            startButton.setOnClickListener(v -> {
                if (checkPermissions()) {
                    if (checkBackgroundPermission()) {
                        startSensorService();
                    } else {
                        requestBackgroundPermission();
                    }
                } else {
                    requestPermissions();
                }
            });
        }

        if (stopButton != null) {
            stopButton.setOnClickListener(v -> stopSensorService());
        }

In this application, when the user taps the start button, the application checks both permissions in sequence, and the stop button stops the service.


checkPermissions() :
This method checks at runtime whether the BODY_SENSORS permission has been granted. On Galaxy Watch, the user must explicitly grant this permission on their device.

private boolean checkPermissions() {
        return ContextCompat.checkSelfPermission(this, Manifest.permission.BODY_SENSORS) == PackageManager.PERMISSION_GRANTED;
    }

checkBackgroundPermission() :
This method checks for the BODY_SENSORS_BACKGROUND permission, which is essential for Wear OS 3+ devices (like Galaxy Watch 5, 6, 7) to access sensor data in all power states.

private boolean checkBackgroundPermission() {
        if (Build.VERSION.SDK_INT >= Build.VERSION_CODES.Q) {
            return ContextCompat.checkSelfPermission(this, Manifest.permission.BODY_SENSORS_BACKGROUND) == PackageManager.PERMISSION_GRANTED;
        }
        return true;
    }

startForegroundService() :
On Android 8 (Oreo) and above, you must call this method instead of startService() when starting a foreground service.

private void startSensorService() {
        Intent intent = new Intent(this, SensorService.class);
        if (Build.VERSION.SDK_INT >= Build.VERSION_CODES.O) {
            startForegroundService(intent);
        } else {
            startService(intent);
        }
        Toast.makeText(this, "Sensor Service Started", Toast.LENGTH_SHORT).show();
    }

stopSensorService() :
Once the task is completed, call this method to reduce battery drain.

private void stopSensorService() {
        Intent intent = new Intent(this, SensorService.class);
        stopService(intent);
        if (heartRateTextView != null) {
            heartRateTextView.setText("--");
        }
        Toast.makeText(this, "Sensor Service Stopped", Toast.LENGTH_SHORT).show();

requestPermissions() :
This method prompts the user for the BODY_SENSORS permission before starting the service.

private void requestPermissions() {
        ActivityCompat.requestPermissions(this, new String[]{Manifest.permission.BODY_SENSORS}, PERMISSION_REQUEST_CODE);
    }

requestBackgroundPermission() :
This method prompts the user for the BODY_SENSORS_BACKGROUND permission. Since the sample application targets Android 13 (API level 33) or higher (currently set to 34), this permission is required if you want to access sensor data in the background, even when using a foreground service. Without it, the system can restrict or stop sensor data delivery when the application is not in the immediate foreground for an extended period.

private void requestBackgroundPermission() {
        if (Build.VERSION.SDK_INT >= Build.VERSION_CODES.Q) {
            Toast.makeText(this, "Please allow 'All the time' sensor access in settings",Toast.LENGTH_LONG).show();
            // On API 30+, background permission MUST be requested separately and 
            // the user must be directed to settings manually in many cases, or through a system dialog.
            ActivityCompat.requestPermissions(this, new String[]{Manifest.permission.BODY_SENSORS_BACKGROUND}, BACKGROUND_PERMISSION_REQUEST_CODE);
        }
    }

Override the onRequestPermissionsResult() method to handle the user's response to each permission request:

@Override
    public void onRequestPermissionsResult(int requestCode, @NonNull String[] permissions, @NonNull int[] grantResults) {
        super.onRequestPermissionsResult(requestCode, permissions, grantResults);
        if (requestCode == PERMISSION_REQUEST_CODE) {
            if (grantResults.length > 0 && grantResults[0] == PackageManager.PERMISSION_GRANTED) {
                if (checkBackgroundPermission()) {
                    startSensorService();
                } else {
                    requestBackgroundPermission();
                }
            } else {
                Toast.makeText(this, "Permission denied to read sensors", Toast.LENGTH_SHORT).show();
            }
        } else if (requestCode == BACKGROUND_PERMISSION_REQUEST_CODE) {
            if (grantResults.length > 0 && grantResults[0] == PackageManager.PERMISSION_GRANTED) {
                startSensorService();
            } else {
                Toast.makeText(this, "Background permission denied. Data collection may stop when app is not in foreground.", Toast.LENGTH_LONG).show();
                // Optionally start service anyway, knowing it might be limited
                startSensorService();
            }
        }
    }

Even if the background permission is denied, the service is started. This allows heart rate collection to continue while the application is visible, though data collection may pause when it moves to the background.

BroadcastReceiver:
Send an intent with the heart rate value to update the UI components in real time. This should be outside of the onCreate() method.

private final BroadcastReceiver heartRateReceiver = new BroadcastReceiver() {
        @Override
        public void onReceive(Context context, Intent intent) {
            if (SensorService.ACTION_HEART_RATE_UPDATE.equals(intent.getAction())) {
                float heartRate = intent.getFloatExtra(SensorService.EXTRA_HEART_RATE, 0);
                if (heartRateTextView != null) {
                    heartRateTextView.setText(String.format(Locale.getDefault(), "%.0f", heartRate));
                }
            }
        }
    };

onResume() :
Register a BroadcastReceiver inside this method to catch updates and display them in a TextView element.

@Override
    protected void onResume() {
        super.onResume();
        IntentFilter filter = new IntentFilter(SensorService.ACTION_HEART_RATE_UPDATE);
        if (Build.VERSION.SDK_INT >= Build.VERSION_CODES.TIRAMISU) {
            registerReceiver(heartRateReceiver, filter, Context.RECEIVER_NOT_EXPORTED);
        } else {
            registerReceiver(heartRateReceiver, filter);
        }
    }

onDestroy() :
This method stops the service when the activity is destroyed, preventing a dangling service.

@Override
    protected void onDestroy() {
        stopSensorService();
        super.onDestroy();
    }

Step 5: Edit SensorService.java

This is the core of the tutorial. SensorService is a foreground service that registers a heart rate sensor listener and acquires a wake lock to keep the CPU active when the screen turns off.


onCreate() :
Initialize the SensorManager instance and request the wake-up sensor. Here, do not use the default sensor. Instead, request the wake-up version:

sensorManager = (SensorManager) getSystemService(Context.SENSOR_SERVICE);
        if (sensorManager != null) {
            // Attempt to get the wake-up version of the sensor
            heartRateSensor = sensorManager.getDefaultSensor(Sensor.TYPE_HEART_RATE, true);
            if (heartRateSensor == null) {
                Log.i(TAG, "Wake-up heart rate sensor not available, falling back to non-wake-up.");
                heartRateSensor = sensorManager.getDefaultSensor(Sensor.TYPE_HEART_RATE);
            }
        }

Standard sensors stop sending data when the screen turns off. The true parameter ensures the sensor can wake up the processor to deliver data even in deep sleep.

onStartCommand() :
Execute the foreground service notification.

Promoting your service to the foreground is mandatory for the tracking to stay alive. This prevents Galaxy Watch from pausing your application after 60 seconds of screen-off time.
Promote Immediately: In onStartCommand(), promote the service to the foreground immediately to satisfy Android’s background limitations.
Build the Notification: Create a persistent notification that informs the user that heart rate tracking is active.
Specify Service Type: Android 10+ requires the FOREGROUND_SERVICE_TYPE_HEALTH type for health sensors.
Register Listener: Register the sensor listener to begin receiving heart rate events.

Check the code here:

 @Override
   public int onStartCommand(Intent intent, int flags, int startId) {
       Notification notification = new NotificationCompat.Builder(this, CHANNEL_ID)
               .setContentTitle(getString(R.string.sensor_notification_title))
               .setContentText(getString(R.string.sensor_notification_text))
               .setSmallIcon(android.R.drawable.ic_menu_info_details)
               .setPriority(NotificationCompat.PRIORITY_DEFAULT)
               .build();

       if (Build.VERSION.SDK_INT >= Build.VERSION_CODES.Q) {
           startForeground(1, notification, ServiceInfo.FOREGROUND_SERVICE_TYPE_HEALTH);
       } else {
           startForeground(1, notification);
       }

       if (heartRateSensor != null) {
           sensorManager.registerListener(this, heartRateSensor, SensorManager.SENSOR_DELAY_UI);
           Log.d(TAG, "Heart rate sensor registered.");
       } else {
           Log.e(TAG, "Heart rate sensor not available.");
       }

       return START_STICKY;
   }

onSensorChanged() :
To process sensor data and broadcast updates, implement this method to handle the actual data.
Capture Value: Extract the heart rate from event.values[0].
Broadcast Result: Send a local broadcast with the heart rate value so your UI components can update in real-time.

Here is the code:

@Override
    public void onSensorChanged(SensorEvent event) {
        if (event.sensor.getType() == Sensor.TYPE_HEART_RATE) {
            float heartRate = event.values[0];
            Log.d(TAG, "_________Heart Rate: " + heartRate);
            
            // Broadcast the result to update the UI
            Intent intent = new Intent(ACTION_HEART_RATE_UPDATE);
            intent.putExtra(EXTRA_HEART_RATE, heartRate);
            intent.setPackage(getPackageName()); // Ensure only this app receives the broadcast
            sendBroadcast(intent);
        }
    }

onDestroy() :
In this method, unregister the sensor listener to prevent excessive battery drain after the user is finished.

@Override
    public void onDestroy() {
        if (sensorManager != null) {
            sensorManager.unregisterListener(this); //Stop sensor
        }
        Log.d(TAG, "Sensor service destroyed and listener unregistered.");
        super.onDestroy();
    }

Step 6: Download the Sample Application

You may download the final projects here:

SensorReadConExample
(556.0 KB) 04/23/2026

Step 7: Run the Sample Application on Galaxy Watch

To run the sample application on a Galaxy Watch:

  1. Connect Galaxy Watch to Android Studio over Wi-Fi.
  2. Run the sample application on your device.
  3. Tap START SENSORS and grant the sensor permission when prompted.
  4. When you see the second prompt (or toast), go to the System Settings > Apps > Permissions > Sensors and select All the time.
  5. Once granted, the data collection continues even if you close the application UI or the watch screen goes dark.
    When the screen turns off, heart rate logs continue in Android Studio Logcat.
  6. Tap STOP SENSORS when you want to stop data collection. This stops the service.

2026-04-10-01-02.png

Figure 2: Sample application output on a real device

2026-04-10-01-03.png

Figure 3: Data collection output in Logcat

In Logcat, filter by the SensorService tag to see the collected heart rate readings. New readings arrive even while the watch screen is off.

Conclusion

Following the steps above, you can build a Galaxy Watch application that collects heart rate data continuously—even when the screen turns off. This same approach applies to other sensors as well, allowing you to read any sensor data continuously in the background.

If you have any questions about or need help with the information in this article, you can reach out to us on the Samsung Developers Forum or contact us through Developer Support.

View the full blog at its source



  • Replies 0
  • Created
  • Last Reply

Top Posters In This Topic

Popular Days

Top Posters In This Topic

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
  • Similar Topics

    • By Samsung Newsroom
      Samsung Supports the Next Wave of Healthcare Innovation at the 2026 Out-of-Pocket Hardware Hackathon
      The future of healthcare is increasingly being shaped by the convergence of connected devices, artificial intelligence, and developer innovation. That's why Samsung was proud to sponsor the 2026 Out-of-Pocket Healthcare Hardware Hackathon, a unique event that brought together healthcare builders, engineers, clinicians, designers, and entrepreneurs to create the next generation of digital health experiences.
      Over an intensive weekend of collaboration and innovation, participating teams explored how wearable technology, health data, and AI can work together to improve health outcomes, enhance patient engagement, and unlock entirely new healthcare experiences.
      Building the Future of Connected Health
      The hackathon challenged teams to develop solutions that leverage real-world health signals, connected devices, and modern AI capabilities. Projects spanned a broad range of healthcare use cases, including:
      • AI-powered health coaching and guidance
      • Personalized wellness experiences
      • Recovery and resilience monitoring
      • Remote patient monitoring
      • Preventive health applications
      • Health data interoperability and integration
      • Intelligent health assistants and copilots
      Many teams explored how continuous biometric monitoring can enable more proactive and personalized healthcare journeys, highlighting the growing role wearable technology plays in the future of health.
      Samsung's Role in the Developer Ecosystem
      As a sponsor, Samsung provided participants with access to wearable technology and insights into the Samsung Health ecosystem. Several teams leveraged Samsung devices and health data capabilities while building their solutions, demonstrating the versatility of wearable-generated health insights across a variety of applications.
      Beyond supporting the event, the hackathon provided an opportunity to engage directly with developers and innovators, gaining valuable feedback on:
      • Developer onboarding experiences
      • Health data accessibility
      • API and SDK usability
      • Interoperability requirements
      • Emerging healthcare application trends
      These insights help us continue improving the tools and resources available to developers building on Samsung Health technologies.
      Key Trends Observed
      One theme emerged consistently throughout the event: healthcare is becoming increasingly AI-native.
      Participants demonstrated how AI can transform raw sensor data into meaningful, actionable insights that help individuals better understand and manage their health. The combination of continuous sensing, intelligent software, and personalized recommendations was evident across many of the top projects.
      Another notable trend was the growing demand for open, interoperable health ecosystems that allow developers to create seamless experiences across devices, platforms, and healthcare workflows.
      Looking Ahead
      The creativity, technical excellence, and healthcare impact demonstrated during the hackathon reinforce the enormous opportunity ahead for developers building in digital health.
      Samsung remains committed to supporting developers, researchers, startups, and healthcare innovators who are leveraging connected health technologies to create meaningful solutions for consumers and patients worldwide.
      We are excited to continue expanding the Samsung Health developer ecosystem and look forward to seeing the next generation of applications that transform health and wellness through technology.
      To all participants, mentors, organizers, and sponsors, thank you for helping advance the future of connected healthcare.
      Partnering with Samsung Digital Health
      The future of healthcare will be shaped by collaboration. Technology can provide the tools. But partnerships give those tools meaning. At Samsung Digital Health, we’re bringing together technology, research, and care to make health and wellness more personal, and accessible.
      We invite innovators, researchers, and organizations around the world to join us in this work. Together we can build a connected ecosystem where every insight, every partnership, and every idea move healthcare forward, one collaboration, one community, and one user at a time.
      Interested in partnering with Samsung Digital Health? Visit our Partnerships page to learn about our global partner network. Review case studies and documentation to see how existing collaborations work and how we can work together to create and align a partnership.
      View the full blog at its source
    • By Samsung Newsroom
      Samsung Research Poland (SRPOL) is thrilled to announce a transformative collaboration with the Polish National Institute of Cardiology (IKARD). This partnership focuses on enhancing cardiac health research through the use of Samsung Health Research Stack. By integrating advanced technology with medical expertise, this initiative aims to revolutionize the continuous monitoring of patients' health and vital signs, ultimately contributing to more effective prevention strategies.
      A key aspect of this collaboration is the tailored adaptation of Samsung Health Research Stack by SRPOL. By providing necessary adjustments to meet specific business criteria, the solution now offers enhanced capabilities for examining critical health metrics such as heart rate and bioelectrical impedance analysis (BIA). This integration allows researchers to continuously monitor patient data, enabling proactive and personalized healthcare interventions. The focus on continuous monitoring ensures that potential health issues can be addressed promptly, reducing the risk of complications and improving overall patient well-being.
      This partnership highlights the shared commitment of Samsung Research Poland and IKARD to advancing medical research and technology. By leveraging the customized implementation of Samsung Health Research Stack, this collaboration is poised to make significant contributions to the field of cardiac care. Watch our exclusive video to learn more about this groundbreaking initiative and witness the fusion of technology and healthcare innovation in action.
      Join us as we explore the future of cardiac research and care, where continuous monitoring and prevention are at the forefront of improving patient outcomes.
      View the full blog at its source
    • By Samsung Newsroom
      Samsung Health Data SDK has been updated with support for additional health data types, expanding the range of health insights available to developers building digital health apps.
      This update introduces support for two new data areas generated by Samsung Health Monitor on compatible Galaxy Watch devices. It includes accessing Irregular Heart Rhythm Notification (IHRN) and sleep apnea data.
      Irregular Heart Rhythm Notification (IHRN)
      The SDK provides a new data type to retrieve Irregular Heart Rhythm events detected on Galaxy Watches by Samsung Health Monitor application. While the app itself can notify users when potential irregular rhythms are detected, thanks to Samsung Health Data SDK you can now access history record of all these events. Having such data, you could easily observe heart rhythm patterns and elevate monitoring of your health condition.
      Sleep apnea data
      Support has also been added for sleep apnea data reported by Samsung Health Monitor. This feature analyzes overnight data collected across multiple nights to assess signs associated with obstructive sleep apnea. Developers can query these results through the SDK to incorporate sleep health insights into their apps.
      These data types are intended for informational and integration purposes only and are not a substitute for medical diagnosis or treatment.
      With these additions, developers can leverage the Samsung Health Data SDK’s existing APIs for data reading and change queries to work with a broader set of cardiovascular and sleep health data, enabling richer healthcare experiences.
      You’ll find these new features in Samsung Health Data SDK. Refer to Release Note for more information.
      View the full blog at its source
    • By Samsung Newsroom
      Samsung introduced an innovative Electrodermal Activity (EDA) sensor in the Galaxy Watch8, offering detailed insights into stress and sleep patterns. This technology enables users to better understand their well-being by analyzing physiological responses.
      The EDA sensor provides precise monitoring of electrodermal activity, which is crucial for assessing stress and sleep quality. With Samsung Health Sensor SDK, developers can create innovative solutions that leverage Galaxy Watch8 raw data, supporting users in health monitoring and sleep improvement.
      EDA measurement is available with the Galaxy Watch8 series.
      At the World Sleep Congress 2025 in Singapore, the potential of EDA technology was demonstrated, showcasing its role in advancing digital health.
      Explore our code lab for hands-on guidance on implementing EDA data in your applications. To find more information, see Measure Electrodermal Activity on Galaxy Watch.
      View the full blog at its source
    • Government UFO Files
    • By Samsung Newsroom
      October 2025 Move to Samsung Health Data SDK as Samsung Health SDK for Android Deprecates
      Samsung Health SDK for Android was deprecated on July 31, 2025. The SDK will remain available for an additional two years and will reach the end of its service in 2028. To ensure continuity and gain access to additional health data types, please migrate to the Samsung Health Data SDK as soon as possible. Migrating to the Samsung Health Data SDK will allow you to continue using existing data types and access new ones. For detailed instructions for the migration, please refer to the Migration Guide. SmartThings Q3 Updates: Next-Gen Hub, Thread, & Analytics
      SmartThings introduced significant new products for users and developers in Q3. Highlights include two-way thread unification, and enhanced analytics providing WWST partners with deeper insights into product usage. The next-generation Aeotec Smart Home Hub 2 offers faster performance with double the RAM, enabling it to connect to more devices with support for Matter, Thread, and Zigbee. The WWST ecosystem also continues to grow, incorporating more brands and products. Read more to discover the key SmartThings updates in Q3 here. Samsung Health Accessory SDK v3.2 Released
      The latest version of Samsung Health Accessory SDK, version 3.2, now supports integration with indoor bikes and cross trainers, enhancing the fitness experience for users. Additionally, the Verification Tool’s testing scope has been expanded to include these devices, offering a more reliable testing environment. Utilize the upgraded SDK today to create innovative health solutions. Explore more features and detailed information here. Deprecation of Tizen Studio
      The Tizen Studio will be deprecated with the next SDK release (Platform 10, SDK 6.5). The primary IDE for Tizen application development will transition to the Tizen Extension for Visual Studio Code. This change streamlines development, leverages Visual Studio Code's robust ecosystem, and introduces enhanced features for a smoother, more efficient Tizen application development experience. Resources will be provided to ensure a seamless transition. Discover more about this change here. Easy Steps to Capture and Collect Logs from Galaxy Watch
      Have you used the dumpstate log on Galaxy Watch? It's a comprehensive system diagnostic report that captures everything happening on the watch at a specific moment. Developers can utilize it to identify issues such as application crashes, excessive battery drain, and sensor failures that cannot be diagnosed from application logs alone. It is also useful for deep debugging, capturing everything from application behavior to hardware status in one file. This tutorial demonstrates the step-by-step process for collecting the log, whether or not your smartphone is connected to a Galaxy Watch. Generate the dumpstate log on your connected smartphone, wearable applications, and Galaxy Watch to more quickly and accurately assess the watch's status.

      Explore more​​​​​​​ Create Samsung Wallet Card Templates Using the Server API
      Samsung Wallet partners can create and update card templates to meet their business needs through the Wallet Partners Portal. However, managing a large number of cards on the website can be challenging. Samsung offers server APIs that allow partners to easily create and modify Samsung Wallet card templates without accessing the Wallet Partners Portal. This tutorial guides you through creating a card template for coupons using the Add Wallet Card Templates API. The API enables an expanded card management via an independent UI or dashboard, with secure and flexible integration through a base URL, authentication headers, and a structured request body based on JWT. Discover how to create new card templates on the server rather than the portal, enhancing operational efficiency for your wallet on our blog.

      Learn more Samsung’s Breakthrough Wearable Technologies Driven by Innovation and Collaboration
      Samsung Electronics has developed the world’s first smartwatch feature to detect heart failures early, which received approval from the Ministry of Food and Drug Safety in September. The AI algorithm for early detection of LVSD, developed in partnership with the Korean medical device company Medical AI, has proven reliable in real-world use, now serving more than 100,000 people each month across over 100 hospitals worldwide. Samsung also recently collaborated with the Department of Biomedical Engineering at Hanyang University to create an around-the-ear wearable device capable of measuring brain waves (EEG), advancing innovation in brain-computer interface (BCI) technology. The ergonomically designed device detects high-quality signals through electrodes around the ear and has demonstrated real-world application potential, such as drowsiness detection and video preference analysis. Explore how Samsung is shaping the future of healthcare through constant innovation and collaboration with leading experts and institutions on our Newsroom.

      Read more Clustering-based Hard Negative Sampling for Supervised Contrastive Speaker Verification
      In recent years, researchers have employed various deep learning approaches to enhance the performance of speaker verification (SV) systems. Contrastive learning methods for SV are gaining attention as alternatives to traditional classification-based approaches because they effectively utilize hard negative pairs, which are different-class samples that are particularly challenging for a verification model to distinguish due to their similarity. 

      Samsung R&D Institute Poland introduces a clustering-based hard negative sampling (CHNS) method to improve the efficiency of supervised contrastive learning models for SV. The CHNS approach clusters embeddings of similar speakers and adjusts batch composition to obtain an optimal ratio of hard and easy negatives during contrastive loss calculation. Learn more about how CHNS enables high-performance SV even in environments with limited computing capabilities on the Samsung Research blog.

      Learn more SemEval-2025 Task 8: LLM Ensemble Methods for QA over Tabular Data
      Large language models (LLMs) have shown exceptional capabilities in understanding and generating natural language, making them widely used in question answering (QA) tasks. However, they still encounter difficulties when processing and reasoning over tabular data, particularly in understanding relationships, identifying relevant columns, and answering complex queries.

      To tackle these challenges, the Samsung Research team presented a new tabular QA solution at the SemEval-2025 Task 8 competition. This solution is based on an ensemble approach using generative LLMs, where each model contributes to question interpretation, column selection, code generation, and result verification. The final answer is produced through a voting mechanism. The team also introduced automated pandas/SQL code generation with iterative correction and cross-verification between LLMs, significantly improving both accuracy and reliability. The solution achieved 86.21% accuracy and ranked second amongst 100 teams participating in the competition. This new tabular QA solution helps LLMs better understand and utilize structured data, paving the way for broader applications such as automated data analysis, AI assistants, and search systems. Discover more about this innovation on the Samsung Research blog.

      Learn more View the full blog at its source





×
×
  • Create New...