Image Processing, Biomedical Processing and Robotics
13
Scopus Publications
Scopus Publications
SmartYatra: a dynamic driver profiling and vehicle condition monitoring safety shield for accident prevention Sri Lasya Koti, Abhishek Koti, Pallavi Khare, Akhil Khare Urban Planning and Transport Research, 2025 This paper presents SmartYatra, an intelligent road-safety framework that integrates driver-behaviour profiling and vehicle-condition monitoring to reduce the risk of accidents. Road accidents are a major concern worldwide, claiming countless lives and leaving countless others injured, and a significant number of them are due to unsafe and reckless driving behaviours combined with poor vehicle conditions. This paper introduces an intelligent road-safety system aimed at reducing such road accidents dynamically by profiling the driver’s behaviour and monitoring the vehicle's condition in real time, which are then used to regulate the maximum allowable speeds tailored to the vehicle in use, thereby ensuring safe driving on the roads. The system also provides instant alerts when there are unsafe driving events and generates weekly performance reports that summarise driver behaviour and vehicle-condition scores, thereby promoting safer driving habits and proactive vehicle maintenance. The results obtained from a simulated environment demonstrated the effectiveness of the system in reducing unsafe driving habits and events, as well as in enhancing road-safety metrics. The accuracies of around 90% were achieved in driver profiling, and experiments benchmarked against rule-based telematics and intelligent speed assistance (ISA) show up to a 78% reduction in overspeed exposure for aggressive profiles (≈52% for normal profiles), achieved using comfort-aware advisory limits, with an overall improvement of 17% in unsafe-event detection and 8% in adaptive-speed accuracy compared with these baseline methods; absolute cruising speeds remain unaffected when compliant. It employs a weighted-scoring algorithm that first computes a driver behaviour score (Bd) from sensors and a vehicle condition score (Vc) from simulated diagnostic parameters. These are fused into a safety score (Sc = αBd + βVc), which dynamically determines the maximum safe speed (Ms) for each trip segment. The design is edge-first, offline-capable, and low-bandwidth, enabling real-time advisory control on commodity devices with minimal back-haul. It establishes a foundation for further improvement and development of intelligent transport systems, such as driver fatigue detection and autonomous vehicle compatibility. This system therefore represents a significant step toward minimising fatal accidents on the road and in maximising and fostering a safer driving environment.
Dual-Mode Currency Identifier and Travel Assistance Using AssistEye Srilasya Koti, Abhishek Koti, Pallavi Khare, Akhil Khare Proceedings of the 2025 12th International Conference on Computing for Sustainable Global Development Indiacom 2025, 2025 This study presents AssistEye (dual-mode currency identifier and travel assistance), which is an initiative in assistive technology that is designed to assist visually impaired individuals in public transportation systems and in currency recognition. The visually impaired individuals face many challenges in identifying currency denominations using just touch and reliance on others. The public navigation systems are also heavily dependent on people with good vision. The proposed assistive solution, AssistEye is a compact, affordable and offline assistive electronic device that can be handheld and portable. The device combines image processing for currency note identification, and QR (Quick Response) code scanning for real-time audio-based travel assistance. A camera module of the device captures the currency images, and the several QR codes intended to be placed at public transport entrance points provide the transport route and schedule information. The methodology involves integrating the hardware components including Raspberry Pi, a camera module, QR scanner, and a text-to-speech engine, along with software modules for image processing using OpenCV and QR code decoding. The conducted experiment using the proposed solution successfully achieved 95 % currency recognition accuracy and 98% QR Code success rate under optimal conditions. This indicates it to be a practical alternative to expensive assistive technologies and internet-dependent solutions. This project can further work on global currency recognition and provide with multilingual support for a diverse user demographic in the future, along with advanced navigation features using GPS (Global Positioning System) and public transport APIs (Application Programming Interface). AssistEye demonstrates the potential of technology to make a tangible difference in the lives of visually impaired individuals, to enhance their independence and to promote greater inclusivity.
Beyond the Paradigm: Unraveling the Limitations of Object-Oriented Programming Abhishek Koti, Sri Lasya Koti, Akhil Khare, Pallavi Khare Multifaceted Approaches for Data Acquisition Processing and Communication, 2024 This paper examines the limitations of Object-Oriented Programming (OOP), a popular software development paradigm, and its applicability in various contexts. The study uses a literature review to analyze OOP’s evolution and adoption, highlighting potential challenges such as performance overhead, memory consumption, and scalability issues. These issues are more pronounced in high-concurrency environments and large-scale distributed systems. The paper uses quantitative analysis and case studies to support its claims, comparing OOP with other programming paradigms in terms of execution time, CPU usage, and memory efficiency. The results show a noticeable performance trade-off associated with OOP, especially in resource-constrained scenarios. The paper advocates for a balanced approach to selecting programming paradigms, recognizing its limitations and promoting a multi-paradigm approach in software development. It emphasizes the importance of a critical and informed application of OOP, recognizing its limitations, and integrating it judiciously with other paradigms to develop efficient, maintainable, and robust software systems.
Solar-smart hydroponics farming with IoT-based AI controller with mobile app Pallavi Khare, Abhishek Koti, Akhil Khare 2023 IEEE Sustainable Smart Lighting World Conference and Expo Ls18 2023, 2023 Only Hydroponics uses less water and other resources than traditional soil-based farming. Hydroponics using mechanical aid. Controversies with hydroponics. In the realm of gardening, hydroponics is king. Water, temperature, and pH are all controlled by it. Typical household current is used. We seldom see a spike in electricity use. The investigation resulted in two potential courses of action. Hydroponics with environmental temperature control. Hydroponics solar power plant second. Light intensity, battery voltage, DC-to-AC converter current, and solar panel voltage are all tracked by the smart solar power plant. Server-side telemetry of water storage tank’s temperature, acidity, depth, brightness, and power supply. In this research, a solar-powered hydroponic automation system was created, allowing for precise regulation of acidity/ alkalinity/ temperature/ water level/light/current/voltage. Hydroponics farming is difficult to monitor because of the many parameters, nutrition, and plant diagnosis systems. Recent technological advances help solve these problems by adopting AI-based controlling algorithms in agriculture. The work focuses on implementing a mobile application integrated AI-smart hydroponics system with IoT. The proposed smart hydroponic system with IoT consists of three phases.
A Digital Toolkit designed to assist those with Visual Impairments Abhishek Koti, Akhil Khare, Pallavi Khare 2023 IEEE Sustainable Smart Lighting World Conference and Expo Ls18 2023, 2023 Only 25% of what we know about our surroundings comes from our other senses, while 75% comes from our vision. Patients with BVI are mobile both indoors and outdoors. The instruments benefit both parties. Despite their efforts, they did not meet user and technological standards. Scientists from a variety of disciplines investigate how people with BVI think and feel about space. This is required for the BVI patient investigation. Complicated are networked sensing, computing outsourcing, internal placement, and release. “Cognitive mapping,” an extensively researched concept, can aid the mobility of blind or visually impaired individuals. This facilitated the use of systems by blind and visually impaired individuals. This study calls for additional research into how blind or visually impaired individuals utilise clever technology. Blind individuals use Braille to read and write. Even PCs and smartphones require modifications. The quality of existence will improve as a result of the introduction of new instruments. Modern devices use optical character recognition and text-to-speech to convert written text into audible speech. Write or print. Every device has defects. A impaired person’s third sight is used to create hardware and software. The third eye restores vision. The third vision was designed for the blind. Both are complete. This is the effect of ultrasonic waves generated by a sensor or alarm. The individual who utilises it can distinguish it from other items. This concept was inspired by the long-distance trekking canes used by blind individuals. Weak individuals must use canes.
Computational Framework for the Sustainable Management and Optimization of E-waste Proceedings of the 17th Indiacom 2023 10th International Conference on Computing for Sustainable Global Development Indiacom 2023, 2023
Location and Time Aware Resource Seeking Framework for Mobile P2P and Ad Hoc Networks Akhil Khare, G Madhu, Pallavi Khare Proceedings of the 2022 9th International Conference on Computing for Sustainable Global Development Indiacom 2022, 2022 A Mobile Ad hoc Network (MANET) is a network with low infrastructure that allows users to create a mobile, wireless network dynamically. Due to the similar nature of its dispersed components, P2P overlay networks and mobile wireless ad hoc networks have several fundamental properties such as self-organization, dynamism, resilience to failure, and decentralization. It's critical to develop P2P algorithms that can handle the issues provided by mobile ad hoc networks. P2P MANETs can be used when communication infrastructure is unavailable due to natural disasters or political turmoil. It's critical to develop P2P algorithms that can handle the issues provided by mobile ad hoc networks. The major challenges of MANET are found in these networks, including node mobility, resource constrained nodes, and content transaction with anonymous nodes. It becomes necessary for the P2P network to observe the change that are taking place in the underlying MANET to form the groups of nodes based on content similarity. Most earlier studies focused on location-based peer clustering but neglected to account for time when adopting resource search strategies, resulting in inefficient search results.
