International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 10 Issue: 05 | May 2023 www.irjet.net p-ISSN: 2395-0072 © 2023, IRJET | Impact Factor value: 8.226 | ISO 9001:2008 Certified Journal | Page 1489 “A DEVICE FOR AUTOMATIC DETECTION OF ELDERLY FALLS” Falgun Padme 1 , Vitthal Biradar 1 , Jay Kulkarni 1 , Prof. P.P. Gaikwad 2 1 Student, Department of Electronics and Telecommunication, Sinhgad College of Engineering, Pune, Maharashtra, India 2 Assistant Professor, Department of Electronics and Telecommunication, Sinhgad College of Engineering, Pune, Maharashtra, India ---------------------------------------------------------------------***--------------------------------------------------------------------- Abstract - Falls by elderly individuals and patients could be dangerous if not caught in time. The idea is to create a fall detection system that, in the event of an emergency, sends an SMS to the involved parties or to the doctor. Continuous monitoring of patients who are unwell and prone to falling is required to reduce falls and the harm they cause. The suggested solution involves creating a prototype of an electronic device that is used to detect falls in older people and those who are at risk for them. In this article, the change in acceleration in three axes—measured using an accelerometer—is used to determine the body position. To measure the tilt angle, the sensor is positioned on the lumbar area. To minimise false alarms, the acceleration values for each axis are compared twice with a threshold and a 20- second delay between comparisons. The threshold voltage values are chosen using experimental techniques. Microcontrollers are used to carry out the algorithm. The GPS receiver, which is configured to track the subject continually, pinpoints the position of the fall. When a fall is detected, the gadget communicates by sending a text message via a GSM modem Key Words: Fall Detection, GPS , GSM , Accelerometer. 1. INTRODUCTION Falls are a primary gamble component of injury for old matured individuals and it is a critical boundary to seniors' free living. They are a main source of injury-related hospitalizations in individuals who matured 65 years or more. As indicated by the past factual results, somewhere around 33% of individuals matured 65 and up fall at least one times each year [1]. After a fall episode happened, a harmed old individual might be left on the ground for a few hours or even days. Habitually, the individual probably won't have the option to ascend with no help or on the other hand support and could require quick clinical thought. Likewise, there is a reality that, feeling of dread toward fall is produced or connected with the fall occasion. So particularly for senior individuals who have encountered falls before, most certainly will tend to stay away from doing everyday proactive tasks. It makes a pessimistic sensation of weakness to them assuming nobody is there. For forestalling the serious results of this fall, persistent or consistent fall identification is required. Human fall discovery framework notice and arranges everyday life exercises of human to distinguish an accidental fall. To distinguish human falls, different sensors and procedures have been utilized to characterize everyday exercises. Specialists have arranged fall location frameworks into three classifications in light of cameras, wearable gadgets, and feeling sensors. Among the wearable gadgets, accelerometer is the most generally utilized strategy to understand a fall. It utilizes the proportion of the speed increase of the body to characterize falls. Clifford et al protected a human body fall location framework utilizing accelerometer, a processor, and a remote transmitter. The processor utilizes accelerometer measurements to decide whether the individual with wearing the gadget is falling and there is a non-development stage followed by the fall. The created reaction is then, at that point, somewhat sent to a transmission beneficiary by a remote transmitter [2]. Research are being attempted to decide human fall utilizing the stance developments. Body direction as stance development is utilized to distinguish a fall utilizing either pose sensors or different accelerometers. Kaluza et al introduced a stance-based fall location calculation utilizing the philosophy of reconstruction of an article's stance. The stance reproduced in a 3D plane by finding the remote labels which were put on body parts (sewn on garments, for example, shoulders, lower legs knees, wrists elbows and hips. Some labels are additionally positioned at explicit positions like bed, seat, couch, table to recognize a few stances, for example, lying on bed or sitting on seat. The fall location calculations use speed increase edges alongside speed profiles. Speed increase is gotten from the developments of the labels. Speed increase and precise speed computation is dependent upon the label's confinement accuracy [3]. Kangas et al utilized a midriff worn tri-hub accelerometer, handset, and microcontroller to foster another fall identifier model in light of fall related effect and end pose [4]. Afterward, Li et al introduced a clever fall location framework utilizing both accelerometer and spinners. By utilizing two tri-pivotal accelerometers at isolated body areas they can perceive four sorts of static stances: standing, twisting, sitting and lying. Movements between these static stances are thought of as powerful advances and if the progress prior to lying stance is not deliberate, a fall is distinguished. Whether movement changes are deliberate or not set in stone by the straight speed increase and rakish speed estimations [5]