Journal of Innovations in Data Science and Big Data Management

Soft Computing: Emerging Diversifications for Innovation and Sustainable Practices

2026-02-26T06:24:57+00:00

Soft computing is a paradigm of biologically inspired computational techniques that effectively tackle complex, real-world problems involving uncertainty, imprecision, and partial truth. Soft computing contrasts with conventional “hard computing” by prioritizing approximate, robust, and low-cost solutions over precise, exact ones. Its core components including fuzzy logic for reasoning with vagueness, artificial neural networks for learning and pattern recognition, and evolutionary computation for optimization are often integrated into hybrid systems to leverage their complementary strengths. In technical literature, the abstract of an Adaptive Neuro-Fuzzy Inference System (ANFIS) paper typically highlights its role as a hybrid model that merges the transparent, human-like reasoning of fuzzy logic with the data-driven learning of artificial neural networks. Applications of soft computing are widespread across various fields, including healthcare (medical image analysis), finance (fraud detection), and manufacturing (industrial robotics). While soft computing has seen tremendous success, challenges remain in model interpretability, real-time scalability for resource-constrained environments like edge computing, and robustness against adversarial inputs. Continued research and development in hybrid systems, explainable XAI, and optimization techniques for resource-efficient deployment are crucial for soft computing's future, ensuring its continued relevance in the ever-evolving landscape of artificial intelligence. By addressing these challenges, soft computing can continue to drive technological innovation and support sustainable advancements that benefit society.

Runtime Safety‑shielded Deep Reinforcement Learning Approaches for Collision Avoidance in Autonomous Vehicles

2026-02-26T06:59:58+00:00

Self-driving technology is leaning heavily on deep reinforcement learning (DRL) to handle the complex, ever-changing nature of traffic. While DRL has shown it can be effective at avoiding accidents, there is a catch: it rarely comes with a solid safety guarantee once the car is actually on the road. The problem is that most current methods focus on safety only during the training phase. This leaves the system vulnerable if it encounters a scenario it has not encountered before or if conditions shift too rapidly. Obviously, that is a major roadblock for using these systems in real cars where safety is non-negotiable. This paper tackles that exact issue by introducing a “safety shield” framework designed for real-time collision avoidance. Instead of just hoping the AI remembers its training, this system runs alongside the learned policy while the vehicle is moving. Because this safety layer is separate from the learning process, it makes the entire system more reliable without requiring a rebuild of the core architecture. This approach in simulated traffic is tested, and the difference is clear. The shielded framework significantly cut down on crashes and safety breaches compared to standard DRL, all while keeping up with the driving task just as efficiently. These findings suggest that adding an explicit safety check during operation is not just helpful it is a practical, necessary step to getting DRL-based cars safely onto real-world streets.

Smart Grow: IoT-based Real-time Plant Care Assistant with Telegram Alerts

2026-03-23T10:28:37+00:00

Smart Grow is a low-cost internet of things (IoT)-based plant-care assistant designed to simplify indoor plant maintenance through continuous environmental monitoring and instant user communication. The system employs an ESP32 microcontroller integrated with a capacitive soil‑moisture sensor and a BH1750 digital light sensor to measure two critical parameters affecting plant growth: water availability and light intensity. Sensor readings are processed locally, filtered, and compared against predefined thresholds to determine plant health conditions in real time. Whenever abnormal conditions such as dry soil or insufficient illumination are detected, the system sends immediate notifications and care suggestions to the user through a Telegram bot interface. Unlike conventional monitoring dashboards, the messaging‑based interaction allows users to receive updates, request live readings, and customize thresholds without technical expertise. Experimental evaluation demonstrated stable measurements, rapid alert delivery, and reliable Wi‑Fi reconnection during interruptions. The modular architecture enables future expansion with temperature, humidity, and automated irrigation modules. Overall, Smart Grow provides an accessible and scalable solution for students, households, and beginner gardeners by combining sensing, wireless communication, and user‑friendly alerts to minimize plant neglect and promote sustainable plant care practices.

A Comprehensive Survey on AI-enabled Timetable Scheduling for Multidisciplinary Education under NEP 2020

2026-04-03T07:29:01+00:00

With a focus on flexibility, interdisciplinarity, and learner choice, the National Education Policy (NEP) 2020 has complicated the scheduling of academic programs. Historically, administrators have relied on computerisation for academic timetable generation, examples of scheduling, elective coursework and faculty preferences. As a result, institutions may struggle to create conflict- free schedules, classroom utilisation, and balance faculty workload as a result of administrative scheduling problems. This paper outlines the development of an academic Artificial Intelligence timetable generation system (TGS) using heuristic optimisation and constraint satisfaction methods to automate and optimise the scheduling process. The system is capable of processing a range of institutional data inputs such as course structures, faculty availability, classroom sizes, and student enrollments. It consists of three primary functional modules—ADIL (Academic Data Integration Layer), ISL (Intelligent Scheduling Logic), and AOL (Adaptive Optimisation Layer)—that automatically communicate and provide a process to develop efficient, flexible, and conflict-free schedules. The ADIL gathers and normalises scheduling data from a variety of academic departments, the ISL uses heuristic algorithms and rule-based logic to dynamically apply constraints, and the AOL further optimises the schedule through a combination of adaptive optimisation strategies to improve overall efficiency. An experimental evaluation conducted in a university environment demonstrates that the proposed system achieves a substantial reduction in manual scheduling effort and a 70 percent improvement in timetable generation time. Additionally, the system enhances conflict resolution accuracy, scalability, and adaptability to policy- driven academic frameworks.

Multiscript Handwritten Receipt Recognition and Information Extraction Using Transformer Architectures

2026-04-03T09:16:24+00:00

The digitization of handwritten receipts written in Indian languages presents challenges due to script diversity, handwriting variability, and irregular document layouts. Conventional OCR systems, primarily optimized for printed or English-centric data, often fail to generalize effectively to Indic scripts containing conjunct characters and diacritical modifiers. This study proposes an end-to-end framework for multilingual handwritten receipt recognition and structured transaction extraction. The system integrates a transformer-based OCR model for script-aware text recognition with a semantic processing layer for contextual interpretation of extracted content. Preprocessing techniques are applied to enhance visual clarity under degraded imaging conditions, while schema-guided language modeling converts unstructured OCR output into structured financial records. The framework also supports natural language-based transaction queries for improved usability. Experimental evaluation using character error rate (CER), word error rate (WER), and transaction extraction accuracy demonstrates improved robustness over baseline OCR systems. The proposed solution provides an integrated approach for intelligent receipt digitization in multilingual environments.