The Smart Phone is Mostly a Sensor Platform

Posted on July 14th, 2014 by admin in Scenarios, Technology

If you only causally pay attention to business  discussions about smart phones by pundits’ and listen to their speculations about the next iPhone, you could easily start to think it’s all about watching videos and social media. You would miss seeing the new reality of the smart phone as a sensor platform.

The ubiquitous mobile phone is emerging as one of the most important sources of data about people and their environment. Every time someone uses a cell phone, information can be collected, stored and analyzed, including the phone’s time stamped geo-location, characteristics of the user’s voice patterns deduced by signal processing on-phone or by operators in the cloud, imagery captured by the user, medical emissions and physical activity. This enormous range of possible applications is enabled by a variety of measurements taken by sensors on the phone.

Only 10% or so of hardware on a smart phone is dedicated to phone calls per se. The remainder is all about control of displays, battery, compute-memory, file storage and managing sensors. The cell phone generally is becoming a platform to support major emerging applications.

We are most interested in medical applications of smart phones. All medical data derived from various cell phone sensors usually require professional evaluation, and thus the necessity for a network connection to upload the data for evaluation (the phone part). Further, many mobile applications exist for the convenience of the user to avoid an office visit  (e.g., analysis of a blood sample), or to enable real-time measurements (e.g., heart rate monitoring), or to become elements of public health systems.

In the public health context, APPs are designed to aggregate many users’ sensor data to form large scale socio-medical conclusions. For example, a cloud based program could evaluate multiple readings for people with high body temperatures to infer a flu epidemic may be starting.

Basic sensors

The sensors listed in Table 1 provide basic knowledge about the phone’s overall external state and the sensors all provide meaningful information to the user. These sensor outputs are listed in the Android Developer’s Kit and the software instructions to use this data in applications are defined there. These are standard measurements.

Table 1: Basic Sensors

Sensor Device Core measurement
gyroscope, hand set orientation
accelerometer, change in velocity (2 or 3 axes)
magnetometer, magnetic field strength
barometer, ambient air pressure
thermometer air temperatures.
proximity sensor, distance to screen of object
ambient light sensor brightness
GPS receiver geo-location of hand set
camera records optical field
gravity motion detection

Examples of how some of these measurements are provided by commercial phones are listed in Table 2 for Apple and Samsung.

Table 2: Examples for Apple and Samsung

iPhone: 2007 Galaxy S: 2010 iPhone 5S: 2013 Galaxy S5: 2014
Accelerometer x x x x
Proximity x x x x
Ambient light x x x x
3-axis gyro     x x
Fingerprint     x x
Compass       x
Barometer       x
Gesture       x
Heart rate       x

For basic sensors built-in to the phone, some advanced applications have been developed and are always improving.  See Table 3.

Table 3: Applications based on sensor data

Face detection and recognition
Gesture recognition
Palm tracking
Augmented reality
Object detection and tracking
Emotion detection

To provide real specific medical applications, new sensors will need to be embedded in the smart phone.

Table 4: Future Sensors and Medical APPs

Medical Sensor Add-Ons
Heart rate
Breathing rate
Blood pressure
Blood sample analysis (Glucose/CO2/NOx/etc)
Geiger counter/Dosimeters

 

While all these sensors are embedded devices in the cell phone and make specific physical measurements, when they are used in combination or in non-intuitive ways, one can get dramatic new applications. For example, consider the work of Prof. Ki Chon, a biomedical engineer at Worcester Polytechnic Institute (WPI) in Massachusetts. He developed an Android APP that measures not only heart rate, but also heart rhythm, respiration rate and blood oxygen saturation – all through a finger against the camera lens. Measurements made by the app are said to be as accurate as those obtained using standard medical monitors. He’s now working to  adapt the APP to recognize atrial fibrillation, the most common form of cardiac arrhythmia. See this link for a news report with more detail.

 

Leave a Reply

More News