The concept of “Smart Chair” was the requirement of many engineering field corporations to know the basic and ideal sitting posture and also the ideal conditions of the spinal cord and lumbar and hip support to last its help for comfort of humans for longer time. Many intellectuals did a pretty well research in this line particularly, some of them are mentioned below:
In the late 1800s and early 1900s, several design items were identified. Very early, Parow and von Meyer concluded that the ischial tuberosities were the chief points of support in the sitting position because of posterior pelvic rotation. Von Meyer stated that spinal ligaments were not in tension while sitting and that support is required to give lumbar relaxation. He noted that straight-back chairs did not give support to the spinal column. In 1884, Staffel’s designed chair had lumbar support and space under this support for the buttocks to slide backward to affect some forward rotation of the pelvis. Here in Fig. 3, the hip support and related bones are justified by the center of mass of the body and the inclination.
After knowing much about what was required for sitting, consulting physiotherapists, came to know that there can be different problems that can arise after sitting in 4 different types of positions which are mentioned in the introduction. In 1953, Keegan noted that seat-bottom height, when too high above the floor, could cause shorter people to have dangling legs. This position causes compression stresses on the soft tissues of the posterior thigh and becomes uncomfortable in a short period. Thus, a short person sitting on a chair that is too high for him or her will soon sit on the edge of the chair and negate any seat-bottom incline or any lumbar support.
The concept of the smart chair resembles Yue Li and Rachid Aissaoui’s smart wheelchair which can analyze the trunk flexion angle and pelvic rotation angle along with the pressure while the person is sitting on it. Using linear regression model and Network proposal, they tried to monitor and detect the position of the person. The model measures the sitting position only, not to update the user anything like to change the position and the data was in quantitative and it was not recorded for future analysis [4].
Mayer and the team proposed a chair with a textile pressure sensor for the seating posture classification. Here electrodes are built on both sides of conductive textile material. This structure forms a variable capacitor. Based on different seating postures it creates different pressure systems on the textile pressure sensor. That changes the capacitance of the circuit. It is one of the pioneering works toward detecting seating posture and opens up a new field to monitor human posture for medical treatment. However, due to the limited advancement of IoT and cloud technology at this time, this work does not contain a mechanism for online monitoring and detection system for wrong sitting postures [5].
Jingyuan Cheng and his colleagues did a fantastic job of detecting different positions with an accuracy of 88.06% for 5 subjects and 7 classes, just by placing the pressure sensor at the bottom of the chair’s leg. Also, found and detected the subtle hand and head actions like typing, etc. Combining both, they got the real-time data of 4 subjects and 7 classes with an accuracy of 78.3%. They have not used any cloud technology to show real-time data to user which was the limitation of this project [6].
According to Byeong-Gu Ahn and co-authors says that people want to live healthy and immortal life and IoT has been playing an indispensable role in making the health care system much more advanced like Smartphone-based real-time health management solutions. They have studied how postures change affects the bio signal’s measurement of person’s heartbeats as he/she changes the position. The only limitation of this study was heart rate can be influenced by certain emotions and it may cause an error in reading. Also there is no feedback loop to notify the user for bad posture [7].
Furthermore, an low power consuming Smart chair was made by J. Paek and their team, which was based on separated seating pads so that, sensors on each pad can record the data from a particular data and the program can identify the posture using the Arduino, iBeacon, and Bluetooth network to commute the data from sensor to the smartphone. The sensor used in this study were 6 in number and can read accurate data. Even though, the data of the sensor was not shared via the cloud and no cloud technology is been used. It is for low consumption of power and is not beneficial for long distance communication [8]. The similar product was designed and implemented by Ganesh, Srinu and the team, using the Bluetooth data is been transmitted to the custom android application and then the screen shots and data are sent to the doctor via Email or MMS for quick feedback which was the improvement of the [9] design. The main intention to make this was the reduction of cost but the range and the use of Arduino increased the hardware complexity which was not favorable.
To know whether this chair will improve the sitting posture or not, C. C. Roossien, with the help of pears, got to know that monitoring the sitting position and giving feedback at regular intervals, a 12-week prospective cohort study was done among office workers (Approx. 45) and was regularly given the notification and it was noted that, After turning off the feedback signal, a slight increase in sitting duration was observed (10 min, p = 0.04), a slight decrease in optimally supported posture (2.8%, p < 0.01), and musculoskeletal discomfort (0.8, p < 0.01) was observed. We conclude that the ‘smart’ chair is able to monitor the sitting behavior, the feedback signal, however, led to small or insignificant changes [10].
Firstly, an automated attendance system was created which was used in schools and universities which help in manual attendance. Using the IoT and from the generated database a student-teacher interaction portal was created to make the institute smarter. But from the database obtained it was found that students had problems in the postures they were sitting on which would, later on, generate complications like Endometriosis, Sacroiliac Joint Dysfunction, Degenerative Disc Disease, and more. So later on, they added a feature regarding improper sitting and to maintaining a good healthy lifestyle [11].
Along with all this, George Flutur, with his fabulous team innovated a chair with a sensor embedded on it, and flags from the microcontroller are being used as the notification to the user. It measures the back features and detect the posture which harms the back of the person. It only focused on the back inclination problem of wrong sitting posture[12].
Hu et al. proposed a Field Programmable Gate Array (FGAP) based smart chair equipped with a two-layer Artificial Neural Network (ANN). They use only flex sensors with Spartan-6 FPGA kit. The proposed work provides a good level of accuracy but the notification mechanism is not up to the mark. Moreover using the FPGA kit and ANN makes it a complex and costly system [13].
Advancement in the field of Machine Learning and Artificial Intelligence has been increasing day by day. Muhammad Usman and his partners had done a great job of designing a smart chair. They used a pressure sensor for recording the data when a person sits on the chair. They identify the 10 different trails of humans sitting by extracted 6 features. They use the Naïve Bayes and Multilayers preceptor and SVM algorithms for which accuracy stands as 98.75% for two objects, 80% for 3 activities and 98.75% using Naïve Bayes classifier as shown in Fig. 4 [14].
As discussed in the introduction, during the covid-19 outbreak, everyone was forced to stay away from each other. Mr. Waseem Hussain and two other co-authors had done a great job to make a chair to maintain the social distancing feature which can be added to our project just to follow the guidance given by the government [15]. Also, initially, the smart chair was used to measure the pulse and heartbeat of the customers while they were sitting on a chair, which was innovated by R. Suganya and the other 4 teammates. In addition to that, it monitored the sitting positions using node MCU and Bluetooth 4.0 and a smartphone application was created to accept and detect the information on various seats. This chair using IoT is developed which helps customers to sit in a correct posture [16].
As per the above discussion, many authors have given significant contributions in the field of the smart chair. Based on the above study, this work proposes IoT-enabled smart chair to detect 4 wrong postures with minimum sensors and connected to the cloud. It provides a holistic way of the notification mechanism. Moreover, cloud technology helps to create a database of the seating behavior of a person, which helps to find the root cause of medical problems like spinal injury, problems in leg joints, etc... Section 1 of the paper contains basic need for a smart chair and a literature review. The methodology of the proposed works has been explained in section 2, followed by the result discussion and conclusion in sections 3 and 4 respectively.