Coordinator (Black Box) End Device (Wearable Device)

Transcription

1 2017 IEEE 7th International Advance Computing Conference WIRELESS DETECTION SYSTEM FOR HEALTH AND MILITARY APPLICATION Yallalinga* School of ECE Reva Institute of Technology and Management Bengaluru , India Nirmalkumar S. Benni** School of ECE Reva Institute of Technology and Management Bengaluru , India Sunilkumar S. Manvi*** School of Computing and IT REVA University Bengaluru , India Abstract This study is to develop a wireless detection system for health and military application. Among the elderly population falling/collapsing is a common and very serious issue. It can cause severe injuries like fractures, joint dislocations, head injuries etc. In cases where person is immobilized or lost consciousness it is not possible to seek immediate help or medical attention. There are several cases where elderly people live alone are not found for hours after the fall which makes the situation more severe. The damage caused because of failure in seeking medical attention at the right time is more than the damage and injuries caused by the fall itself. Similar cases can be found with Border Security Force soldiers and other security people standing for long hours guarding the nation borders. Also VIP securities standing in front of VIP homes are vulnerable for such attacks causing injuries. Objective of this paper is to develop a wireless fall-detection system which can sense such situations and alerts others seeking help. The sensors for fall Detection is wearable by the potentialvictim/user. Upon detection of fall/collapse the sensor system transmits the information wirelessly, which will be received by the care-taker s mobile. The sensor is a belt shaped wearable device consisting of accelerometer (tri-axial) and gyroscope. These sensors are used to classify the posture and dynamics of the user. The main aim of the project is to develop efficient algorithms to detect falls and distinguish between falls and non-falls using these sensors. The sensor is a part of zigbee node. It will communicate to the microcontroller over zigbee. Zigbee will provide an efficient wireless communication with the microcontroller with relatively small hardware size. The microcontroller is responsible for sending information to the Care taker s mobile using a GSM module interfaced to it. The microcontroller is interfaced to a PC in order to burn the program on to the microcontroller and for testing and debugging. Keywords Wireless Network; GSM model; Tri-axial accelerometers and Gyroscopes; Fall detection; lower quality of life. I. INTRODUCTION According to the World Health Organization (WHO) approximately 28-35% of people aged 65 and over fall each year increasing to 32-42% for those over 70 years of age. The frequency of falls increases with age and frailty level. In fact, falls exponentially increase with age-related biological changes, which is leading to a high incidence of falls and fall related injuries in the ageing societies. If preventive measures are not taken in the immediate future, the number of injuries caused by falls is projected to be a 100% higher in In this context, assistive devices that could help to alleviate this major health problem are a social necessity. Indeed, fall detectors are being actively investigated. A wireless detection machine can be defined as an assistive device whose essential objective is to alert whilst a fall occasion has took place. In an actual-lifestyles state of affairs, they've the ability to mitigate some of the adverse outcomes of a fall. Specifically, wireless fall detectors can have an immediate impact on the discount within the worry of falling and the rapid provision of assistance after a fall. In reality, falls and fear of falling depend upon every different an individual who falls may also sooner or later expand worry of falling and, vice versa, the worry of falling may boom the threat of laid low with a fall. Fear of falling has been proven to be associated with terrible consequences together with avoidance of sports, much less bodily hobby, falling, despair, reduced social contact and lower quality of life. Therefore, many fall detection systems have been developed. However, prevalent methods only use accelerometers to isolate falls from activities of daily living (ADL). This makes it difficult to distinguish real falls from certain fall-like activities such as sitting down quickly and jumping, resulting in many false positives. Body orientation is also used as a means of detecting falls, but it is not very useful when the ending position is not horizontal, e.g. falls happen on stairs /17 $ IEEE 161 DOI /IACC

2 Human activities are divided into two categories: static postures and dynamic transitions. By using tri-axial accelerometers and gyroscopes, four kinds of static postures are recognized they are: standing, bending, sitting, and lying. Motions between these static postures are considered as dynamic transitions. To determine whether motion transitions are intentional or not for that linear acceleration and angular velocity are measured. The block diagram of the wireless detection system (End device) is shown in the figure 1.1 If the transition before a lying posture is not intentional, a fall transition is detected. This algorithm is joined with accelerometers and gyroscopes, reduces both false positives and false negatives, for improving fall detection accuracy. Further via this solution features low computational price and actual-time reaction 2.2 Coordinator Device Fig.1.1 End Device Block Diagram The main block diagram of the wireless detection system (coordinator) is shown in the figure 2.1 Fig. 2.1 Coordinator Block Diagram II. METHODOLOGY Accelerometers and gyroscopes are made used to detect the fall. All the systems designed till date use accelerometers, but for the purpose of accuracy this solution includes a gyroscope also Hence this solution makes use of a MPU 6050 sensor. 2.1 Essential Devices Coordinator (Black Box) Wireles End Device (Wearable Device) Fig 2.1 Basic Block Diagram Two most important devices in our project are as shown in the figure 2.1. CO-ORDINATOR END DEVICE The coordinator end is a black box with power supply that can be placed at any corner of the house which houses the main operation of the Health Monitoring System. Similarly, the end device functions by detecting the fall from the wearable sensors and transmits the data to the coordinator to send emergency messages to the terminal contact (attender in this case). Fig 2.2 Coordinator Device Coordinator device is as shown in the figure 2.2. A Coordinator is a stationary device. It consists of the following: CC2430 Hardware module Maestro 100 GSM Module 3 LEDs to define operations Buzzer Power Supply Magnetic Base antenna Main functionality and the working of Coordinator device can be explained as follows: The CC2430 has 8051 microcontroller and Zigbee module integrated into the same device. This device is used as the main control unit to interface the GSM modem and to send the messages to the emergency contact and also to raise an alarm to the attender if in the vicinity of the patient using a Buzzer (PORT 1.1). The microcontroller has an in-built regulator that regulates the power supply from the 9V adapter or the battery to 3.3V. There are 3 LEDs which are used basically to represent the status of the various conditions at the Coordinator end. These conditions help the user of the device to easily setup the device. Green LED (PORT 1.2) indicates the Power ON condition. The Red LED (PORT 1.0) indicates the battery low condition at the Coordinator end when battery is used to power up the device. The Blue LED (PORT 2.0) indicates the status of the communication link between Coordinator and the end device. The GSM communicates with the CC2430 via RS 232 serial interface using UART protocol. The GSM is interface using the serial port present in CC2430. There are USART registers present in CC2430 which is set to UART mode. 162

3 The Port 0 Pin 2 and Port 0 Pin 3 are used for the purpose of transmission and reception respectively. Transmission is done by setting up the various registers such as U0CSR to the function in UART mode, setting up the port as a general purpose input or output and also setting the baud rate at One of the other register used during both transmission and reception are the transfer and receive flags which are set and cleared accordingly to perform the respective operations. By reception we mean that the device checks the information sent back by the GSM after the command messages are being sent to it. The instructions being sent from CC2430 to GSM are AT commands. AT commands are the summon set comprises of a progression of short content strings which join together to deliver complete orders for operations, for example, dialing, hanging up, and changing the parameters of the connection.. In our case these set of AT commands are specifically used for the purpose of sending messages of various cases to the emergency contact stored in the coordinator device. The flow developed for the communication between GSM and CC2430 is to first initialize the device. The second step in this process to test the communication by checking the SIM card status. The next step in this process is to send the message to stored emergency contact. As and when the conditions of fall detected or battery low at the coordinator end or at end device, sensor failure or end device out of range is detected, related functions are invoked to send the message. Whenever a data of FF arrives from the end device to the coordinator, the GSM sends a message FALL DETECTED to the contact. When a data of BB arrives a message END DEVICE BATTERY LOW is sent to the contact. The message COORDINATOR BATTERY LOW is sent according to the threshold value which is 2.5V. The coordinator device will check for a time period of 4 seconds whether there is a link between itself and the end device to invoke a END DEVICE OUT OF RANGE message. 2.3 End Device The end device block diagram is shown in figure 2.2. This is the wearable part of the project. The end device has been made into a compact unit consisting of the sensor node and the microcontroller [1]. The sensor used is MPU-6050 manufactured by Invensense. The CC2430 microcontroller module is used to control the logic of the end device. I/O ports are used to interface the microcontroller with the other devices such as sensor and LEDs. The microcontroller module consists of an in-built oscillator producing clock frequencies of 32MHz. A 2.4GHz antenna is used to communicate with the coordinator end with radio frequencies using Zigbee protocol. The MPU-6050 sensor mainly consists of an accelerometer and a gyroscope. The accelerometer is used to measures linear acceleration and tilt angle. The 3-axis accelerometer detects the combined magnitude and direction of linear and gravitational acceleration. It can be used for providing limited motion sensing functionality. For example, a device with an accelerometer can distinguish development from a vertical to horizontal state in a settled area. Accordingly, accelerometers are fundamentally utilized for detecting device introduction as for gravity. The gyroscope is used to measures the angular rate of rotation about one or more axes. The gyroscope can quantify complex movements precisely in free space, thus making it a required movement sensor for following the position and revolution of a moving object. Unlike the compasses and accelerometer, gyroscopes are not depends on any external forces such as gravity and magnetic fields, and can therefore function fairly autonomously. The gyroscope measures 1degree angular rate which corresponds to the value 131. An array of LEDs is provided on the face of the end device to provide indications to the user. The three LEDs are for power on, radio connection establishment and battery low. The LEDs are connected to the port0, pin0 for power on, pin1 for connection establishment and pin2 for battery low of the CC ohm resistors are provided to control the LEDs connected through active low logic. The device is operated by two AA 1.5V batteries. The end device can be worn around the neck as a necklace or strapped onto the chest. 2.4 Algorithm Fig 2.3 End device Fig 2.4 Flowchart of Algorithm 163

4 The working of the Health Monitoring System is as follows: 1. The coordinator broadcasts for 10ms to establish a link with an end device (or a peer). 2. If the end device is not available, then it means that communication between coordinator and end device is not established. a. The coordinator now enters into sleep mode and sleeps for 2s and if the communication is established after sometime then it goes to STEP 1. b. But if the communication is not established even after sometime, then a message Patient out of range will be sent to the emergency contact. c. Also a buzzer will be triggered during this and STEP 1 will be repeated. 3. But if the end device is available, the BLUE LED at both the ends (coordinator and end device) will start blinking indicating the establishment of communication link. 4. Now, there are two possible cases here: a. The patient is in normal condition ( i.e., no fall has occurred or he is involved in a ADL activity) i. The coordinator again goes into sleep mode to save power. ii. And after waking up, keeps monitoring to see if any change has occurred. If no change continues monitoring else goes to next step. b. The patient is in abnormal condition (i.e., a fall has occurred) i. If the fall is detected, End device continuously transmits a message to the coordinator end and the buzzer will be triggered at the coordinator end. The coordinator will send an emergency message to the terminal contact. ii. If the fall is not detected, There are two possible reasons for this: Sensor failure End device out of range Either way, a corresponding emergency message will be sent to the terminal contact via the coordinator and a buzzer will be triggered. III. RESULTS AND DISCUSTION Case-1. End device Battery low: CC2430 has an inbuilt ADC, which is used to detect battery low condition. A resolution of 12 bits is set. Whenever the supply voltage reduces below 2.5V, which corresponds to a reading of 1526, the code BB is sent to the co-coordinator device. And red LED is used to indicate the low battery in end device Fig 3.1 Case-2. Patient/Soldier fall Detect: Fall is defined as a sudden transition to horizontal position. The MPU 6050 sends the accelerometer and gyroscope readings to CC2430 using I2C protocol. Thresholds are set for those readings. Threshold readings for accelerometer and gyroscope are 8000 and 5000 respectively. If the readings are above thresholds then fall is detected. After fall is detected the code FF is sent to the co-coordinator end via Zigbee protocol. The process of reading the accelerometer and gyroscope values, and comparing them with the thresholds goes on continuously once a link is established between the end device and cocoordinator. And gsm module will send the message to the mobile and buzzer is also starts beeping. Fig

5 Case-3. Out of Range: The co-coordinator end is linked with the end device through radio frequency using Zigbee protocol. The range of zigbee is 80m line of sight. If the patient crosses this range, cocoordinator end tries to link with end device for 4 seconds (50 times). If the link cannot be established even then, the co-coordinator end sends an out-of-range message to the mobile and buzzer will also start beeping. Fig 3.5 Fig 3.3 Case-4. Sensor Fail: The MPU-6050 contains an algorithm for self-test. Using this feature, the sensor failure condition can be checked. When the end device is turned on, the value in the who_am_i register is checked. If the value is equal to 68, then the sensor has passed the self-test. If the value of who_am_i is not equal to 68, then the code BB is sent via the radio frequency using Zigbee protocol to the co-coordinator. Fig 3.4 Case-5. Coordinator device Battery low: CC2430 has an inbuilt ADC, which is used to detect battery low condition. A resolution of 12 bits is set. Whenever the supply voltage reduces below 2.5V, which corresponds to a reading of 1526, the code BB is sent to the co-coordinator device and coordinator send message, and red LED is used to indicate the low battery in coordinator device. IV. APPLICATIONS Enhanced for monitoring the elderly: For their day to day health and safety this system can be used with ease and comfort. Need for a person to always stay with the elderly can be avoided and elderly people can perform their day to day activities without fear. Used in hospitals to monitor patients: Can be used to monitor patients who are required to walk after their treatment. Also the location of the patient can be easily detected by doctors. Used widely for monitoring children during their initial stage of learning to walk. Useful for epileptic patients: In most of the cases, epileptic patients fall unconscious, which may cause injuries too, by use of such systems these events, can be avoided. In Industrial settings it could alert first responders of possible incidents in the event that human reporting is slow or hesitant about reporting accidents. Monitoring of fall detection of soldiers and VIP securities in borders and VIP homes. V. CONCLUSION In this paper, a wireless sensor system is applied to measure the acceleration and angular velocity at center chest on the body for four different types of fall. The collected data is used to evaluate the performance of the wireless fall detection algorithm. There are many fall detections systems investigated in previous studied. Unlike the previous fall detection systems, we have integrated both accelerometer and gyroscope to give accurate results. Use of both these sensors together helped us to differentiate between ADL (Activities of Daily Living) and a fall. Apart from sensing the fall, we have also covered the conditions of battery low, sensor failure and patient out of range. Also the device can be installed without any complications making it easy to be operable for anyone. Also the device consumes less power making it energy efficient. We also have used small battery 165

6 operated devices, which can be easily wearable. Experiments to evaluate detection accuracy are conducted. The experimental results show that our detection system achieves good performance in terms of low false negative and low false positive in fall detections. This system is applicable not only to elderly but also to healthy individuals performing various activities walking, running, climbing, cycling, rolling, etc. experiencing falls due to various causes such as: unexpected health problems, inattention, dangerous environment, car accidents, attacks, etc. With the completion of this project, we have accomplished our aim of making a robust device which will be helpful to the elderly and patients making their life safer. The device offers many advantages and is greatly sophisticated than the previously implemented systems which used the same principle. Research and development in this field will prove to be extremely helpful in the Medical fraternity and Electronics industry. The project setup was designed keeping the end user in mind making it ready to be implemented in a consumer specific environment. VI. FUTURE SCOPE Wireless fall detectors want to be as accurate and dependable as feasible. A strong wireless fall detection system has to showcase each high sensitivity and specificity. That is occasionally reached in experimental environments, but whilst implemented to an actual state of affairs, the detection price decreases. Those devices are designed and tested underneath managed conditions, for example they use information from falls and ADL of younger human beings simulated on the discretion of every creator due to the dearth of a standardized process or a public database /.l3q233edd comparison. Furthermore, it's far worth pointing out that fall detectors are geared toward older people, so that they need to additionally be involved of their development. simplest few research include records from older humans, although their participation is restricted to perform a hard and fast of simulated sports of day by day dwelling for a few minutes or hours. That isn't enough to evaluate the device overall performance in a real state of affairs. Users need to wear the devices for longer durations (at least months). A few studies have labored on this course ensuing in a considerable wide variety of fake activations, amongst different issues. A wireless fall detection system has been validated to show high sensitivity and specificity results, using the combination of accelerometer and gyroscope. This system can be made even more accurate by using devices such as magnetometer and DMP. The main task ahead is not only to give a sensitive device but also miniaturize the size of the wearable device by integrating both MPU 6050 and Microcontroller onto a single chip platform. This chip can be designed to include various LED s and buzzer for ease. The end device consists of a whip antenna which can be replaced by integrating a chip dielectric antenna of dimension mm which provides the same bandwidth and range. Future development will investigate a system to include GPS and GPRS to inform medical attention. Also these systems can be used in hospitals and old age homes by simply including range extenders to cover large areas. Furthermore, an algorithm that can distinguish ADL such as walking, standing, sitting down/ standing up, running, and jumping will also be investigated. REFERENCES [1] J. Chen,K. Kwong, D. Chang, J. Luk, and R. Bajcsy Wearable sensors for reliable fall detection Shanghai, China, Sept 2005 IEEE. [2] Kangas, A. Konttila, I. Winblad, and T. Jams Determination of simple thresholds for accelerometrybased parameters for fall detection Lyon, France, Aug 2007 IEEE. [3] Qiang Li, John A. Stankovic, Mark Hanson, Adam Barth, John Lach Accurate, Fast Fall Detection Using Gyroscopes and Accelerometer-Derived Posture Information University Of Virginia. [4] Stefano Abbate, Marco Avvenuti, Paolo Corsini and AlessioVecchio Monitoring of human movements for fall detection and activities recognition in elderly care using wireless sensor network: a survey, Italy. [5]Jin Wang, Zhongqi Zhang, Yuhui Zheng, A Novel Fall Activity Recognition Method for Wireless Sensor Networks, School of computer & software, Nanjing University of information science & technology, Nanjing , China.. [6] Q. Li, J. A. Stankovic, M. A. Hanson, A. T. Barth, J. Lach, G. Zhou, Accurate, fast fall detection using gyroscopes and accelerometer-derived posture information, in: Proceedings of the International Workshop on Wearable and Implantable Body Sensor Networks, IEEE Computer Society, Berkeley, CA, USA, 2009, pp [7] D. Wild, U. Nayak, B. Isaacs, How dangerous are falls in old people at home?, British Medical Journal 282 (1981) 266. [8] S. Abbate, M. Avvenuti, P. Corsini, J. Light, A. Vecchio, Monitoring of Human Movements for Fall Detection and Activities Recognition in Elderly Care Using Wireless SensorNetwork: a Survey, in: G. V. Merret, Y. K. Tan(Eds.), Wireless Sensor Networks: Application-Centric Design, InTech, Rijeka, Croatia, 2010, pp [9] N. Noury, A. Fleury, P. Rumeau, A. Bourke, G. Laighin, V. Rialle, J. Lundy, Fall detection - principles and methods, in: Proceedings of the 29th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBS), pp [10] M. Kangas, I. Vikman, L. Nyberg, R. Korpelainen, J. Lindblom, T. Jms, Comparison of real-life accidental falls in older people with experimental falls in middle-aged test subjects, Gait & Posture 35 (2012) [11] S. Sadigh, A. Reimers, R. Andersson, L. La amme, Falls and fall-related injuries among the elderly: a survey of residential-care facilities in a Swedish municipality, Journal 166