Internet of things applications using Raspberry-Pi: a survey

The internet of things (IoT) is the communication of everything with anything else, with the primary goal of data transfer over a network. Raspberry Pi, a low-cost computer device with minimal energy consumption is employed in IoT applications designed to accomplish many of the same tasks as a normal desktop computer. Raspberry Pi is a quad-core computer with parallel processing capabilities that may be used to speed up computations and processes. The Raspberry Pi is an extremely useful and promising technology that offers portability, parallelism, low cost, and low power consumption, making it ideal for IoT applications. In this article, the authors provide an overview of IoT and Raspberry Pi and research on IoT applications using Raspberry Pi in various fields, including transportation, agriculture, and medicine. This article will outline the details of several research publications on Raspberry Pi-based IoT applications. article under the CC BY-SA license.

The rapid growth of IoT technology is anticipated, especially to recent developments in digital technology. As demonstrated in Figure 1, a varied collection of intelligent IoT applications has been developed. Many of them are widely available, and there is still a lot of room for research in areas like smart agriculture, smart transportation, smart health care, smart homes, smart cities, and so on, all of which improve people's quality of life.

IoT APPLICATIONS USING Raspberry Pi
Researchers have proposed a wide range of intelligent IoT applications. This article provides an overview of the advantages of using Raspberry Pi in various IoT applications. This article also exhibits many related applications in various fields, including transportation, agriculture, and medicine.

Intelligent transportation
A key component of a smart city is smart transportation. The idea of a smart city is becoming more and more viable as IoT technology develops. Figure 3 demonstrates some IoT applications in the transportation field.
Kumar et al. [31] implement a surveillance system based on embedded tools and signal to process in autos to develop a comprehensive driver aid system. The implemented system comprises three interconnected modules: detection of driver tiredness, alcohol concentration, and vehicle collisions, as well as monitoring the driver's physiological condition, which can impact vehicle control. The Raspberry Pi was used as the prototype's brain, accessing user-programmed software algorithms on a storage disk card.
Jabbar et al. [32] developed a parking control system to assist staff and students in quickly locating accessible parking places via a smartphone app. The system contains a global policy and strategy (GPS) module to assist users in locating parking spots using the Blynk App, which searches the internet for available parking spaces. The Raspberry Pi gathers and processes the sensors' data.
Jaiswal et al. [33] construct a placing system using IoT and deep learning algorithms to identify the availability of parking slots. The suggested system develops an intelligent parking system using IoT, cloud platform, and machine learning techniques. The suggested system can intelligently resolve various issues, including analyzing available and busy slots in real-time and identifying multiple items in a parking slot, such as a bike in a card slot.
Herrera-Quintero et al. [34] implement a transportation system prototype that includes internet of Things techniques to aid transportation planning for bus rapid transit (BRT) systems. The model can identify Bluetooth signals from various gadgets passengers use while using the BRT system. Based on this information, the prototype can construct an origin/destiny matrix for many BRT routes. The administrative authorities may utilize this information to plan appropriate transportation for the BRT systems. The hardware comprises mostly two components: a sensor module implemented on a Raspberry Pi and a Bluetooth signal sniffer.
Prabu et al. [35] develop a system that leverages IoT technology to construct an effective Raspberry Pi-based parking management system and multidirectional cameras. This suggested solution will aid in the proper management of parking in congested metropolitan areas. Furthermore, OpenCV technology identifies vehicles and their position sectors. The Raspberry Pi device is powered and connected to an external camera to retrieve video data. Raspberry prepares the video outline information to see any available stopping space. Clients may then book a parking space on the site using the results.
Patil [36] suggested a system for monitoring car pollution and noise, with any vehicle exceeding its threshold value reported to the traffic department and national environmental agencies. The suggested system is low-cost and transportable. The sensors are so sensitive that they may detect gas generated by other vehicles while stalled in traffic, jeopardizing the system's integrity. This problem may be solved by reading the samples over time and then taking action depending on the average value.
Bansal et al. [37] proposed DeepBus, a pothole detection system that uses IoT to detect road surface flaws. End users and local governments could see the location of the discovered potholes on a centrally controlled map. The Raspberry Pi has a GPS module and IoT sensors like an accelerometer and a gyroscope. Accelerometer and Gyroscope sensors are used to identify potholes. Various machine learning models are tested using various performance factors to select the most effective pothole identification model and evaluate DeepBus' performance.
Godavari et al. [38] present a vehicle monitoring and tracking system for tracking the school bus from one location to another using Raspberry Pi. GPS is used to determine the vehicle's current location, GPRS transmits tracking data to the server, and global system for mobile communications (GSM) delivers alarm messages to the vehicle's owner's mobile phone. This system also uses gas leakage and temperature sensors to guarantee the passengers' safety.
Husni et al. [39] proposed an IoT system to track fuel use on an android application. The data from the car's engine is read by onboard diagnostics II (OBDII) and relayed through Bluetooth to the Raspberry Pi, which then sends it over a 3 G connection to the server and saves it. Users can access the data via MQTT via a mobile application, then process it on the server for analysis.
Alluhaidan et al. [40] developed a sensor model to monitor a moving vehicle. Data is created as the vehicle moves and encrypted with the Paillier homomorphic cryptography system (PHCS) algorithm before being sent to the cloud for vehicle tracking. The data collected by the various sensors is saved and tracked. If an abnormal condition occurs while driving, an alarm is sounded, alerting the motorist.

Intelligent agriculture
Technology must be used in agriculture because it is crucial for determining the health of the plants. The right plant disease must be identified to use the appropriate chemical. Figure 4 demonstrates some IoT applications in the agriculture field. Ishak et al. [41] proposed an enhanced organic fertilizer mixer remotely monitoring fertilizer output and providing employees with updates and alarms. Compared to today's automated systems, this designed prototype delivers operating cost reductions almost five times. The device is made up of a load sensor mounted to the static drum's bottom and measures the weight of the mixture before sending the data to a microcontroller. The Raspberry Pi was chosen as the microcontroller. The microcontroller receives the sensor's weight data, analyses it, and sends it to the cloud server through a wireless connection.
Arshad [42] produced a prototype that includes a sensor network, IoT analytics, and a Raspberry Pi to monitor climatic factors in a greenhouse setting and transmit parametric climate data to a gateway. The Raspberry Pi is the brain's central processing unit for monitoring real-time climatic parameter values in a greenhouse setting. In turn, depending on these readings, it compares current threshold values to received values and decides whether or not to modify certain climatic parameters if an imbalance is detected.
Benyezza et al. [43] developed a novel approach to monitor irrigation and maintain the appropriate soil wetness to help plants grow. The suggested system was built on a wireless sensors network in several greenhouse zones. This network uses radio-frequency communication to deliver data from the plant environment to a Raspberry Pi.
Using the nitrogen-phosphorus-potassium (NPK) sensor, Lavanya et al. [44] proposed a system that alerts farmers about lacking important soil nutrients such as nitrogen, phosphorous, and potassium. The Raspberry Pi microcontroller is used. The data collected by the built NPK sensor is transferred to a Google Cloud database for quick retrieval. Fuzzy logic is then used to diagnose nutritional deficiencies from sensed data.
Abioye et al. [45] proposed better monitoring of dynamic factors influencing mustard leaf plant irrigation. In this experimental framework, the IoT platform was employed to monitor using numerous sensors and a weather station based on the internet of Things, which revealed better precision irrigation. The acquired data, including plant photos, was sent for online storage to the Raspberry Pi and displayed on the IoT dashboard. MATLAB's system identification toolkit created a mathematical prediction model defining the link between water flow, water loss, and soil moisture.
Goap et al. [46] present an internet of things irrigation system and a hybrid machine learning-based technique for predicting soil moisture. The suggested method incorporates sensor data from the recent past and weather projected data to predict the soil moisture in the following days. Closed-loop control of the water supply is included in the system, allowing for a fully autonomous irrigation plan. The system prototype is inexpensive because it is built on open standard technology.
Ramli et al. [47] present a dependable, smart farm system with an adaptable mechanism. The network protocols used in the proposed system are LoRaWAN and IEEE 802.11ac. The sensor data is sent via the LoRaWAN protocol, which has a tiny data size and uses very little energy. Video data is sent via the IEEE 802.11ac protocol, which offers a faster data rate than LoRaWAN.

Medicine and healthcare
The Raspberry Pi and IoT combo has produced a brand-new, cutting-edge solution in the healthcare industry. The level of patient care will rise due to the inclusion of IoT features in medical equipment. Figure 5 demonstrates some IoT applications in the medical field. Jesudoss et al. [48] present a monitoring, security, and maintenance system for medicine in an IoT context. The suggested platform includes an intelligent medication security box that updates the medicine room. The security camera records the picture of the person who enters the room, and if anybody enters the room, an SMS alert is sent. The VNC viewer program is wirelessly connected to the medication box. It is an Android application that runs on a Raspberry Pi and provides timely medication updates through a sensor.
Alarcón-Paredes et al. [49] presented a non-invasive blood glucose monitoring system. The model comprises a Raspberry Pi Zero powered by a battery bank, a visible laser beam, and a Pi camera integrated into a glove. The Raspberry Pi Zero collects data for glucose monitoring by capturing a series of photographs of the user's fingertip and generating their histograms.
Moghadas et al. [50] present a system for monitoring the cardiac arrhythmia of patients. The k-nearest neighbor method categorizes and validates the kind of cardiac arrhythmia. Instead of transmitting patient information to the cloud, fog technology is used, which reduces data transmission delays. The Arduino Uno and the AD8232 sensor module were used in the proposed system to launch a web service from the Raspberry Pi, allowing real-time electrocardiogram (ECG) and patient heart rate.
Bhatia et al. [51] present an efficient home-based diabetes monitoring system based on urine. For effective prediction analysis based on the temporal features of urine-based diabetes parameters and quantified in terms of level of diabetic infection and diabetic infection measure, a recurrent neural network has been implemented. The suggested system's visualization efficacy has been improved using a self-organized mapping method. Raspberry Pi was utilized to do real-time health data calculations.
Yacchirema et al. [52] developed a novel method to monitor and advise sleep apnea therapy. An edge node in the fog enables IoT connectivity and interoperability and pre-processing IoT data to identify and respond to events in real-time that may jeopardize the health of the elderly. A generic enabler context broker on the cloud maintains, stores, and injects data into the big data analyzer for processing and analysis.
Lomotey et al. [53] presented a wearable IoT architecture for healthcare services, focusing on data traceability. The authors propose an upgraded Petri Nets service architecture to help transparent data-trace path development, tracking, and possibly detecting medical data breaches to overcome the complexity of mapping and matching device data to users. Extensive testing is carried out in real-world circumstances, and the findings reveal that Petri Net adaptability outperforms alternative distributed network models.