On automatizing recognition of multiple human activities using ultrasonic sensor grid

Pervasive or ubiquitous computing environments have become very common nowadays. The vision of pervasive computing is given by Mark Weiser in 1991. Pervasive computing offers such computing environments in which users are offered their desired services without requiring their dedicated or substantial attention. It supports users to achieve their desired goals/task with minimal user interaction. The applications of pervasive computing can be seen in various sub-domains including mobile and ad hoc networks, wireless sensor networks, location-aware systems, context-aware systems and smart environments (e.g. smart offices and smart homes). Sensing physical phenomena, human activity recognition, understanding and analysis can substantially play its role in realizing smart environments. In recent years, the domain of human activities recognition has got significant attention from researchers due its potential applications in smart environments. In the literature of smart environments, it is envisioned that a system could offer the best services to its users only if it is able to understand and recognize the actions/activities being performed by the user while present in the smart environment. Consequently, in this paper a low-cost system to recognize human activities is proposed. The proof-of-concept implementation of the proposed system is carried out and its usability is also conducted. The usability study results show that the proposed system is user friendly and significantly improves the user experience of smart home users.

The HC-Sr04 ultrasonic sensor (proximity sensor) is a widely used ultrasonic sensor, in addition to its affordable price also because of its easy use and easy installation. Ultrasonic sensors are electronic devices whose ability can convert from electrical energy into mechanical energy in the form of ultrasonic sound waves. HC-SR04 sensor is one of the ultrasonic sensors that is often used to monitor the distance of objects (objects) with sensors. This sensor consists of a series of ultrasonic transmitters called transmitters and ultrasonic receivers called receivers. The distance that can be handled ranges from 2 cm to 400 cm, with a precision level of 0.3 cm. The detection angle that can be handled is no more than 15°. The required current is not more than 2mA and the required voltage is +5V. The number of pins is 4. In this study, the authors used the fuzzy logic method to help classify the level of distance received by the sensor. Fuzzy logic is used because this method is able to determine the classification results of the received values, according to what is needed by the distance measuring device with this ultrasonic sensor.

Baby weight and length measurement system has been designed based on ultrasonic and load cell sensors with data storage using MySQL database and website interface. This research aims to produce a baby weight and length measurement system with microcontroller NodeMCU ESP32. The system is designed with an ultrasonic sensor HC-SR04 and four half-bridge load cells of 50 kg. The ultrasonic sensor works on the piezoelectric and inverse piezoelectric principles, while load cell sensors work on strain gauge principles. An ultrasonic sensor measures body length, while a load cell sensor measures body weight. According to the Ministry of Health of the Republic of Indonesia’s regulations, this system also has an LCD 20×4 to display measurement and nutritional status assessment results. The experiment used four doll objects with different lengths and weights. This research shows that the ultrasonic sensor HC-SR04 used in this research has an average error of 0.494% in length measurement. The load cell sensor and HX711 ADC used in this research has an average error of 0.949% in weight measurement. The designed system automatically assesses the nutritional status and stores the measurement result and nutritional status in the database to be displayed on the website.

Aiming at the problem that the single ultrasonic sensor is not concentrated in energy and it is easy to cause false alarm and low resolution with larger beam, an ultrasonic array sensor based on small beam angle is designed to measure the true height between UAVs and the ground. Firstly, the advantage of the wide measurement angle of the ultrasonic sensor is verified by the simulation of the vertical take-off of UAVs accompanying the posture movement; Secondly, the higher resolution of the ultrasonic array sensor is verified by the simulation of UAVs flying flatly into obstacles; Finally, considering the hysteresis of the measurement data of the ultrasonic array sensor, the integrated navigation information is used to fuse by the Federated filter, and the simulation experiments are carried out under various terrain conditions. The simulation results show that the fusion results of the vertical takeoff phase of UAVs and horizontal flight into sinusoidal terrain are better than those of using ultrasonic array sensors alone.

The development of the shipping industry in Indonesia has continued to increase over the past 10 years due to the sabotage principle. This development can be seen from the increasing number of national vessels. The number of national vessels becomes wider from 6,041 units in 2005 to 24,046 units in 2016. The number of vessels makes monitoring the operational performance of the vessel difficult because each ship has different travel routes. This research made tools that can monitor the performance of ships especially the use of MFO aboard the ship. The sensors used to measure the volume of MFO in the tank on board are HC-SR04 ultrasonic sensors and potentiometer pendulum sensors. The data from the sensor was processed by Arduino Uno microcontroller. This study would compare the performance of ultrasonic sensors and potentiometer pendulum sensors mounted on MFO oil premises. The study was conducted by measuring the position of the ship horizontally, right and left tilted with a slope level of 30 degrees and 45 degrees. The result of this research is that the potentiometer pendulum sensor was better when MFO surface condition was flat with the average of error sensor 1.60%, while HC-SR04 ultrasonic sensor has an average of error for 2.87%. However, on the skewed surface conditions of MFO, the usage of HC - SR04 ultrasonic sensors was better with average of error for 3.21%, while the potentiometer pendulum sensor had an average of error for 8.66%.

In this paper, we present a home-monitoring oriented human activity recognition benchmark database, based on the combination of a color video camera and a depth sensor. Our contributions are two-fold: 1) We have created a publicly releasable human activity video database (i.e., named as RGBD-HuDaAct), which contains synchronized color-depth video streams, for the task of human daily activity recognition. This database aims at encouraging more research efforts on human activity recognition based on multi-modality sensor combination (e.g., color plus depth). 2) Two multi-modality fusion schemes, which naturally combine color and depth information, have been developed from two state-of-the-art feature representation methods for action recognition, i.e., spatio-temporal interest points (STIPs) and motion history images (MHIs). These depth-extended feature representation methods are evaluated comprehensively and superior recognition performances over their uni-modality (e.g., color only) counterparts are demonstrated.

Recognizing human activities non-intrusively has prevailed as a challenging and active area of research. In real life, it is a major requirement for human-centric applications like assisted living for elderly care, health-care and creating a smart home environment etc. Considering that people spend more than 90% (Klepeis et al. in J Exposure Sci Environ Epidemiol 11(3):231, 2001) of their time indoors, a proper indoor activity monitoring system will be helpful to monitor the abnormal behavior of the occupants. Existing approaches have implemented intrusive or invasive methods such as a camera or wearable devices. In this work, we present a non-invasive, non-intrusive sensing technique using an array of heterogeneous ultrasonic sensors for human activity monitoring. The ultrasonic sensors are placed in two separate deployments as sensor grids and in different positions of the door-frame. The proposed system senses a stream of events as the occupants perform different activities categorized as primary, postural and group activities. The primary activities considered are sitting, standing and fall. The postural activities are intermediate transitional states in the primary activities. These activities when performed in groups, are considered as a group activity. Other than activities it can identify different indoor movements, count room occupancy and identify occupants. Based on the collected data, the results show that the proposed system achieves an accuracy of more than 90% for detection of different activities and shows improvement over existing works. The final outcome of this work can be seen as developing the current prototype into smart ceiling panels that can be easily used in the indoor environment for human activity monitoring.

Although some IoT systems are built for simple event control where a sensor signal triggers a corresponding reaction, many events are far more complex, requiring applications to interpret the event using analytical techniques to initiate proper actions. Artificial intelligence of things (AIoT) applies intelligence to the edge and gives devices the ability to understand the data, observe the environment around them, and decide what to do best with minimum human intervention. With the power of AI, AIoT devices are not just messengers feeding information to control centers. They have evolved into intelligent machines capable of performing self-driven analytics and acting independently. A smart environment uses technologies such as wearable devices, IoT, and mobile internet to dynamically access information, connect people, materials and institutions, and then actively manages and responds to the ecosystem's needs in an intelligent manner. In this edited book, the contributors present challenges, technologies, applications and future trends of AIoT in realizing smart and intelligent environments, including frameworks and methodologies for applying AIoT in monitoring devices and environments, tools and practices most applicable to product or service development to solve innovation problems, advanced and innovative techniques, and practical implementations to enhance future smart environment systems. Chapters cover a broad range of applications including smart cities, smart transportation and smart agriculture. This book is a valuable resource for industry and academic researchers, scientists, engineers and advanced students in the fields of ICTs and networking, IoT, AI and machine and deep learning, data science, sensing, robotics, automation and smart technologies and smart environments.

This research has realized the design of a smart green trash box based on the HC-SR04 sensor, which is integrated with Arduino Uno. This research aims to design and build a digital instrumentation system based on Arduino Uno, applying Ultrasonic Sensors to the Green Box Trash System. In addition, this system is equipped with an information display via LCD and speakers so that we can find out the contents of the green trash box. Data retrieval by detecting objects with a distance of less than 10 cm to open the trash and measure the volume in the green trash box. The research results are displayed as a prototype Smart Green Trash Box design. Based on the Hc-Sr04 Sensor Integrated by Arduino Uno. The ultrasonic sensor in the Green Box Trash system works well by measuring the volume and can detect when an object is approaching with a distance of less than 10 cm. Assembling the green box using the HC-SR04 sensor, which is integrated with Arduino Uno functions according to the program created if, in the green box, there is HC-Sr04 1 sensor trash, it measure the volume of trash and then display it through the LCD, and if there is an object approaching the HC-Sr04 sensor 2, then the servo moves so that the green trash box can be opened. The Green Trash box can display trash volume and time on the 16x2 character LCD in real-time with the program uploaded to Arduino as monitoring.


 
 
 
 One of the methods that can be used by Indonesian people in urban farming is the Aeroponic method of planting plants. The aeroponic method has a faster nutrient delivery speed than hydroponics and normal plant cultivation. The main objective of this study was to design an aeroponic monitoring and control system for Pakcoy (Brassica rapa. L) using the blynk application as an interface for remote control. The control system used is NodeMCU V3, DHT22 sensor, HC-SR04 ultrasonic sensor and TDS sensor. Plants in the system use a container box measuring 110 cm x 80 cm with 10 nozzles on the inside of the box attached to a ½ inch pipe, with 24 planting holes on the lid of the box with a spacing of 20 cm. The pump will automatically turn on when the temperature in the box is >25oC and will stop when the temperature is =24oC, or when the humidity is <80%. The results obtained during the 30 days of research showed that the average R2 value of the DHT22 sensor was 0.9864 and 0.9864, the HC-SR04 sensor was 0.996 and the TDS sensor was 0.9899. Observations on the growth of pakcoy plants obtained an average height of 13.3 cm for system plants and 13.4 cm for control plants. The average leaf width in system plants was 4.3 cm and 3.6 cm in control plants. the average number of leaves on system plants is 8 leaves and 9 leaves on control plants. The system plant weight was 5.493 gr and that of the control plant was 2.961 gr. The control system that has been made can work well and can be used in aeroponic cultivation of plants.
 
 
 

Human Activity Recognition (HAR) by employing smart home sensors is a topic that is undergoing intense research in the area of ambient-supported living and the basis of ubiquitous computing in smart environments. Recently, HAR majorly utilizes Deep Learning (DL) method since they employ representation learning methods that could automatically produce maximum factors from raw input datasets resulting from the sensor in the absence of human interference and could detect Hidden Layer (HL) patterns in a dataset. The study presents an Improved Manta-Ray Foraging Optimization with Deep Autoencoder (IMRFO-DAE) model for HAR. The presented algorithm majorly recognizes the diverse types of human activities. The presented IMRFO-DAE model pre-processes the human activity data via a standardization approach to obtain this. Next, the IMRFO-DAE method uses the DAE model to perform the Activity Recognition (AR) process. To advance the AR accomplishment of the DAE model, the IMRFO method is applied as a hyperparameter optimizer. Moreover, the IMRFO model is derived by modifying the initialization of the MRFO model by implementing the chaotic concept. An extensive range of simulations was performed to depict the enhanced efficiency of the IMRFO-DAE approach. The simulation results assured the improved outcomes of the IMRFO-DAE approach compared to existing approaches.

Human activity recognition is concerned with identifying the specific movement of a person based on sensor data. In recent years, many different techniques have been proposed for modelling and recognising human activities, with a specific focus on the development of approaches to classifying human activities using deep learning techniques. The research presented in this paper proposes a fuzzy feature representation approach to represent occupancy sensor data, along with a Convolutional Neural Network Classifier (CNNC) for human activity modelling and recognition. Sensory data gathered from a home environment are converted into occupancy data representing human activities and then fuzzified before being fed as inputs into the CNNC. The learning capability of CNNC allows the model to learn the relationship between the fuzzified inputs and their corresponding output activities during training mode. The relations learned in the trained CNNC model are then used to identify human activity patterns and classify these when the testing dataset is applied. The proposed method is evaluated using a dataset representing activities of daily living for a single user gathered from a real-home environment.

Human activity modeling and recognition using wearable sensors is important in pervasive healthcare, with applications including quantitative assessment of motor function, rehabilitation, and elder care. Previous human activity recognition techniques use a whole-motion model in which continuous sensor streams are divided into windows with a fixed time duration whose length is chosen such that all the relevant information in each activity signal can be extracted from each window. In this paper, we present a statistical motion primitive-based framework for human activity representation and recognition. Our framework is based on Bag-of-Features (BoF), which builds activity models using histograms of primitive symbols. We experimentally validate the effectiveness the BoF-based framework for recognizing nine activity classes and evaluate six factors which impact the performance of the framework. The factors include window size, choices of features, methods to construct motion primitives, motion vocabulary size, weighting schemes of motion primitive assignments, and learning machine kernel functions. Finally, we demonstrate that our statistical BoF-based framework can achieve much better performance compared to a non-statistical string-matching-based approach.

Visual impairment significantly hinders daily activities, necessitating innovative solutions. This study develops and tests a smart walking stick using NodeMCU ESP-8266, HCSR-04 ultrasonic sensor, DFPlayer, and GPS, monitored via the Blynk app. Employing a research and development approach, the device demonstrated 100% accuracy in obstacle detection and location tracking, providing real-time feedback through audio notifications and smartphone monitoring. This smart walking stick enhances mobility and safety for visually impaired users, representing a significant advancement in assistive technology. Highlights: Accurate Obstacle Detection: 100% accuracy with HCSR-04 ultrasonic sensor. Real-time Monitoring: GPS tracking via Blynk app. Enhanced Mobility: Increases safety and independence for visually impaired users. Keywords: Visual impairment, Smart walking stick, Ultrasonic sensor, GPS tracking, IoT

Visually impaired individuals usually depend on assistive devices like white canes, frequently equipped with ultrasonic sensors, for navigation. However, these devices face significant limitations, particularly in detecting specific hazards such as potholes and deep trenches on walkways. This gap in functionality increases the risk of accidents and impedes safe, independent navigation for the visually impaired. The research developed a prototype of a blind assistive system equipped with an array of ultrasonic sensors and a Raspberry Pi integrated with Firebase for loT capabilities. AI models, trained on the collected datasets of road images and ultrasonic sensor readings, were deployed on the Raspberry Pi. Testing in real-world scenarios was conducted to validate the prototype's effectiveness. The results showed that the AI model successfully detected potholes with an accuracy of 93%. The prototype could detect both large and small potholes using ultrasonic sensors and a camera but faced challenges in cases where potholes were filled with water or in complex environments.