Ododok: A Health Monitoring App That Tracks Chewing Count Using AirPods

When AirPods Become Health Monitoring Devices

In an era where wearable devices are increasingly ubiquitous, an app called Ododok offers a remarkably imaginative answer: without purchasing any additional hardware, you can use your existing AirPods to track your chewing behavior in real time while eating.

The app recently launched on Product Hunt, garnering 88 upvotes and ranking 10th for the day. Its concept is simple and direct—turning supported AirPods into a real-time chewing tracker. During meals, you can see your chew count, chewing rhythm, chewing duration, and total meal time, with the ability to review data for each meal afterward.

How Ododok Works: How AirPods Sensors Detect Chewing

Ododok's core innovation lies in the clever repurposing of existing hardware. Modern AirPods (especially models with sensors) feature built-in accelerometers and motion sensors, originally designed to detect wearing status, head posture, and spatial audio positioning.

Specifically, AirPods Pro series include multiple sensors: dual beamforming microphones, inward-facing microphones, skin-detect sensors, motion-detecting accelerometers, and gyroscopes for spatial audio. The accelerometer's sampling rate can reach several hundred hertz, originally designed to detect nod/shake gestures (such as hands-free call answering) and dynamic head tracking. After iOS 14, Apple opened up the CMHeadphoneMotionManager API, allowing third-party developers to access AirPods motion data—this is precisely the technical foundation that makes innovative apps like Ododok possible. The framework provides real-time access to four types of data: user acceleration, gravity vector, rotation rate, and device attitude. Developers can obtain these data streams at up to 100Hz, which is sufficient to capture chewing motion patterns in the 1-3Hz frequency range.

From Sensor Data to Chewing Recognition

When a person chews food, jaw movement produces subtle head vibrations and bone-conducted signals through cranial transmission. The motion sensors inside AirPods can capture these regular, minute motion patterns. Ododok leverages this principle, using algorithms to identify the specific action of chewing from the sensor data stream and count it.

It's worth noting that using in-ear sensors to detect chewing behavior is not an entirely new concept. As early as 2015, academic research teams (such as those at the University of Pittsburgh and the University of Tokyo) explored identifying eating behavior through in-ear accelerometers or microphones. During chewing, the temporomandibular joint typically moves at a frequency of 1-2Hz, with each chew producing bone-conducted vibrations of approximately 0.01-0.1g in amplitude. Although these signals are weak, they exhibit distinct periodic characteristics that enable machine learning algorithms to distinguish them from other head movements (such as walking-induced vibrations at approximately 0.5-2Hz). Ododok's contribution is transforming this academic concept into a consumer-grade product, allowing ordinary users to experience it without specialized equipment.

The greatest advantage of this approach is zero additional cost—users don't need to purchase dedicated oral wearable devices or smart mouthguards. As long as you have a pair of supported AirPods and an iPhone, you can start recording your chewing behavior during meals. This dramatically lowers the barrier to health monitoring.

Why Track Chew Count: The Relationship Between Slow Eating and Health

At first glance, "counting chews" might seem like a somewhat peculiar feature. But from a nutritional and health management perspective, chewing behavior is closely related to multiple health indicators.

Benefits of Thorough Chewing for Digestive Health

Extensive research shows that thorough chewing aids digestion by allowing food to mix fully with saliva, reducing the burden on the gastrointestinal system. At the same time, eating too quickly is often associated with overeating and weight gain. When people eat too fast, the brain hasn't yet received the "fullness" signal before excessive calories have already been consumed.

There's a clear physiological mechanism behind this: satiety hormones (such as leptin, cholecystokinin, and GLP-1) take approximately 15-20 minutes to travel from the gastrointestinal tract to the hypothalamic satiety center. A 2014 meta-analysis published in the American Journal of Clinical Nutrition showed that fast eaters had approximately 115% higher obesity risk. A large Japanese cohort study tracking nearly 60,000 people over 3 years found that those who chewed each bite more than 30 times had BMIs averaging 1.5-2 units lower than fast eaters. Nutritionists generally recommend chewing each bite 20-30 times and spending at least 20 minutes per meal, giving the body sufficient time to produce and transmit satiety signals.

Quantifying Eating Behavior with Data

Ododok's four core metrics—chew count, chewing rhythm, chewing duration, and total meal time—essentially quantify "how you eat." This aligns closely with the increasingly popular concept of "mindful eating." Mindful eating is an extension of mindfulness meditation concepts, emphasizing awareness of food and bodily sensations during eating, rather than mechanical, unconscious consumption. Research shows that mindful eating interventions can effectively reduce emotional eating and binge eating, helping participants lose an average of 3-5 kilograms.

By reviewing post-meal data, users can intuitively see whether they're eating too fast or not chewing enough, gradually cultivating healthier eating habits. This immediate feedback mechanism is similar to how fitness bands track step counts—when behavior is quantified and visualized, people are more motivated to make changes.

Product Positioning and Future Possibilities

Ododok is categorized under Health & Fitness, Wearables, and Apple, precisely targeting the health quantification audience within the Apple ecosystem.

A Lightweight Health Tool Defined by Software

Compared to health devices that require dedicated hardware, Ododok takes a "software-defined hardware" approach. It reinterprets AirPods—already worn daily by hundreds of millions of users—as a health data collection endpoint.

The concept of "Software-Defined Hardware" is accelerating in consumer electronics. Tesla unlocking new vehicle features through OTA updates is a classic example, and in the wearable space, Apple Watch didn't support blood oxygen detection or ECG functionality at initial launch—these were later activated through software updates utilizing existing sensors. The business logic of this model is that the marginal cost of hardware is already sunk at the time of manufacturing, while discovering new uses for sensors through software creates incremental value at zero hardware cost. Ododok is a typical example of this approach within the third-party developer ecosystem.

This mindset is inherently inspiring—in the future, more applications may uncover potential uses for earphone sensors, such as posture monitoring (detecting cervical spine risks from prolonged head-down positions), breathing rhythm detection (assisting meditation and stress management), or even analyzing emotional states through heart rate variability.

Accuracy Still Needs Validation

Of course, for inference-based applications relying on consumer-grade sensors, data accuracy remains a concern. Chewing signals are relatively weak and easily interfered with by other head movements like walking, talking, or nodding. From a signal processing perspective, acceleration signals produced by chewing (approximately 0.01-0.1g) are far smaller than head vibrations caused by walking (which can reach 0.3-0.5g), requiring algorithms to perform accurate classification under low signal-to-noise ratio conditions. Additionally, differences in food hardness (such as chewing nuts versus bread) cause significant variations in signal characteristics, challenging the model's generalization capability.

Currently, the product has only 2 reviews on Product Hunt, and actual user experience and recognition accuracy await more user feedback for validation. For users seeking medical-grade accuracy, such tools are better suited as references for behavioral trends rather than precise clinical measurements.

Conclusion: New Possibilities for AirPods Health Monitoring

Ododok represents an interesting product philosophy: rather than piling on hardware, it deeply mines the sensor potential of existing devices, expanding AirPods from audio devices into health monitoring terminals. For users concerned about eating habits who want to practice slow eating, it provides a low-cost quantification tool.

While the practical value and accuracy of chew counting still need time to validate, this app demonstrates the vast potential of combining consumer electronics sensor fusion with software innovation. In an era where everyone owns a pair of earphones, our everyday devices may harbor many yet-undiscovered health monitoring capabilities. As Apple continues to open more sensor APIs, machine learning models gain stronger on-device inference capabilities, and user demand for health quantification continues to grow, the vision of "earphones as health devices" is moving from the laboratory into daily life.