AN EFFECTIVE ALARMING MODEL FOR DANGER AND ACTIVITY MONITORING USING WEARABLE SENSORS FOR CHILDREN P.Vidyadhar 1 , PG Student, Department of ECE, ASCET, Gudur, Andhra Pradesh, India. K. Dhanunjaya 2, Head of the Department, Department of ECE, ASCET, Gudur, Andhra Pradesh, India. Abstract—This paper presents a child activity recognition approach using a single 3-axis accelerometer and a ultra-sonic sensor which is belted around the waist of the baby so as to prevent him from dangers and getting injured. The accelerometer data is collected and monitored in a computer using IEEE 802.15.4 protocol. In addition to the activity recognition child body temperature can also be monitored at regular time intervals. A fire sensor is also embedded in the proposal so as prevent the baby from fire accidents and a SMS using GSM will be sent to their parents if they are going to be involved in any fire accidents. Child activities are classified into 8 daily activities according to our consideration which are moving left, right, front, back, standing still, climbing up, climbing down, and stopping. The accuracy obtained for every activity is around 90% in any situation using single MEMS ADC sensor and ULTRA-SONIC sensor. Index Terms—Accelerometer, activity classification, activity recognition, baby care, child care. I. INTRODUCTION At presentare getting into some accidents due to lack of personal monitoring in this busy world, usually child start walking between 9 and 16 months, there will bechance of falling fromhigher heights or stairs. As the child learns to climb, they will be at risk of falling from stairs, chairs and beds. Children frequent come across injury due to these accidents unknowingly.Medical research show that these accidents are one of the most common cause of injuries that require medical care, and in some situations non fatal injuries also leads to hospitalization. The main areas these accidents occur are at homes because of lack of parental care. Thus, a new effective alarming model for danger and activity monitoring using triaxial accelerometerfor children is required to prevent child from accidents at homes. These accidents have great effect on growth and development of the child.Accident prevention measures are to be taken effectively. The challenging thing is the classification activities in terms of safety and damage occurring to the child. There are many proposals for the recognition of activity but the challenging task lies with the accurately recognition of activity the child is. A smart sensor network is used in this proposal for rescuing the baby from injuries and some small cracks on the body. Multisensor link has been provided for elderly people and children at home. This approach will give the activity data in a simple and recognizable manner. According to theproposal the human activity is recognized by fusing two highly accurate sensors one which is attached to one of the foot and another sensor to the waist of a baby subject, respectively. Due to the use of multiple sensors robustness of the classification systems has been improved drastically and increases thepartialness of high-level decision making. On the other hand, the ultra-sonic sensor could fail to detect the activities like head motion, body tilt, and hand motion. In addition to that and for the purpose of minimizing the number of sensors worn, it is important to know the capability of a certain position to classify a set of activities. Recently, Attallaet al. Investigated the effects of sensor position and feature selection on activity classification tasksusing accelerometers. Accelerometers are the mostbroadly used sensors to recognizeambulation activities such as walking and running the advantage of the accelerometer is inexpensive, require relatively low power, and are greatest applications in most of mobile phones. The study of the optimal sensor concluded that positions depend on the activities being performed by the subject and these activity recognition can be made optimized using accelerometers.The dominant role in the designing the system is that it must be designed in a small space and low weight which can be bearable by the baby. As we are optimizing the sensor activity so we have to compromise in the accuracy. It is difficult to configure an optimal system. It depends on mainly on two important factors one is sensor activity and the positions of the baby. In our study, to decrease the incomfortness the waist belt is kept diaper for the children below three years of age, during physical activity and to measure body motions such as moving left, right forward backward, climbing up and climbing down.In our proposal mainly we have designed a wearable sensor device and a monitoring application to collect information about the activities and to recognize baby activities baby is doing. We classified baby activities into 8 daily activities which are moving left, right, front, back, standing still, climbing up, climbing down, and stopping. As multiple sensors are embedded in a wearable devicewhich are more accurate for collecting differenttypes of sensing informationbut would be very inconvenient for users. TABLE I SENSOR AND TYPES OFTHECOMPONANTS USED TYPE SENSOR VALUE FEATURE Space RFID Room identification Location (kitchen, dining room, bed) Object RFID ID Object Name (electric socket) Activity 3-axis accelerometer [-2g, +2g] Activity (moving left, right) Height Ultra-sonic sensor [30kPa,129kPa] Height from the ground Temperature LM35 [20◦C,100◦C] Ambient temperature Fire Fire sensor - Detection of fire near the baby SMS GSM - Alarming the baby is in danger For this reason, we present only one single unit of sensor nodes, which collects multiple types of information. The nature of information interaction involved in sensor fusion can be classified as competitive, complementary, and cooperative fusion. In competitive fusion, each sensor provides equivalent information about the process being monitored. In complementary fusion, sensors do not depend on each other directly, as each sensor captures different aspects of the physical process. The measured information is merged to form a more complete picture of the phenomenon. Cooperative fusion of the two sensors enables recognition of the activity that could not be detected by each single sensor. Due to the compounding effect, the accuracy and reliability of cooperative fusion is sensitive to inaccuracies in all simple sensor components used. In this paper, we select the cooperative fusion model to combine information from sensors to capture data with improved reliability, precision, fault tolerance, and reasoning power to a degree that is beyond the capacity of each Proceedings of International Conference on Developments in Engineering Research ISBN NO : 378 - 26 - 13840 - 9 www.iaetsd.in INTERNATIONAL ASSOCIATION OF ENGINEERING & TECHNOLOGY FOR SKILL DEVELOPMENT 45