Journal of Microprocessor and Microcontroller Research

Design and Performance Evaluation of a Low-Cost Arduino-Based Automatic Floor Cleaning Robot with Multi-Sensor Obstacle Avoidance

Shreyas Awati — 2026-10-08

This study describes the design, fabrication, and laboratory evaluation of a compact, low-cost floor cleaning robot built around an Arduino Uno (ATmega328P) microcontroller. The machine combines three cleaning actions — dry sweeping with a roller brush, vacuum suction, and a wet mopping pass fed by a small peristaltic pump — and it navigates cluttered indoor spaces using three HC-SR04 ultrasonic sensors together with infrared cliff detectors. A simple reflex-based control algorithm handles obstacle avoidance, edge detection, and tank monitoring without requiring any map of the environment. A bench prototype was assembled for roughly USD 68 in parts and tested on five common floor surfaces in a 25 m² instrumented room. In repeated trials, the robot removed an average of 96.4% of a standard dust mixture from ceramic tile in a single dry pass, responded to obstacles in 0.62 s on average (well inside the 1.0 s safety budget), and covered about 98% of the test area in 45 minutes while consuming 66% of its battery charge. The results suggest that carefully integrated commodity hardware can deliver cleaning performance that approaches entry-level commercial units at a fraction of their cost. All measured data reported here are illustrative results from a bench prototype and are intended as a model experimental framework.

Real-Time IoT-Based 2-Node Electrical Parameter Measurement and Centralized Visualization System

Ankita Hanmant Jadhav — 2026-10-16

Real-time monitoring of energy consumption is important for reducing energy wastage and identifying abnormal electrical conditions at an early stage. Conventional energy monitoring methods often require manual observation and provide limited information about energy usage. This research presents an IoT-Based 2-Node Electrical Parameter Measurement and Centralized Visualization System for continuous monitoring of electrical parameters. The proposed system measures voltage, current, power, energy consumption, and temperature using sensors and processes the acquired data through a microcontroller. The measured parameters are displayed locally on a TFT display and transmitted through an IoT module to a centralized monitoring unit, enabling electrical parameters from multiple locations to be monitored at a single point. This multi-node approach provides real-time measurement, visualization, and automatic data transmission, reducing the need for manual monitoring. The experimental results demonstrate that the system can continuously monitor electrical usage and identify abnormal variations such as sudden changes and overload conditions. The proposed system can be applied in educational institutions, laboratories, industrial facilities, offices, and other energy-consuming environments. Future development can include cloud-based data storage, historical energy analysis, automated fault alerts, advanced energy-management features, and expansion to additional monitoring nodes for larger-scale energy monitoring applications.

AI-Enabled Smart Surveillance System for Anti-Spoofing Using Raspberry Pi, Dashboard, and Simulink

Divya M — 2026-10-27

Facial recognition is now embedded in a wide range of applications, from unlocking personal devices to guarding sensitive infrastructure, which has pushed anti-spoofing research to the front of the security agenda. Conventional surveillance pipelines frequently fall short once an attacker moves beyond a simple printed photo to more advanced presentation attacks, leaving a real security gap. This review looks at how artificial intelligence can be paired with low-cost embedded hardware, specifically Raspberry Pi and MATLAB Simulink, to build surveillance systems that resist spoofing while running at the edge. Attention is given to techniques that exploit edge computing, deep architectures such as convolutional neural networks (CNNs) and You Only Look Once (YOLO), and multi-modal sensing, all aimed at raising both the accuracy and the real-time responsiveness of face anti-spoofing (FAS). The strengths and shortcomings of current methods are weighed against one another, along with how well each copes with 2D print attacks, replayed video, and 3D masks. The review also surfaces the open research problems that remain, including cross-dataset generalization and the computational limits of resource-constrained hardware, with the aim of pointing researchers and practitioners toward more secure, efficient, and scalable AI-enabled surveillance in the future.