The generation of biomedical waste has increased many times after the SARS Cov2 commencement. The biomedical waste generated from COVID-19 Patients is very infectious and contaminated. Thus, it is a big challenge with all stakeholders to avoid spreading of COVID-19 through it. This requires monitoring the complete cycle to the grave to be monitored from the cradle, if the spreading needs to be controlled. The COVID-19 waste generation, collection, storage, transportation and disposal is a big challenge withall stakeholders including isolation wards, quarantine centres, sample collection centres, laboratories, urban local bodies, and the Common Bio-medical Waste Treatment Facility (CBWTF) respectively. As its a novel virus and WHO has instructed that proper guidelines need to be followed with regards to COVID-19biomedical waste generation and its safe disposal. The Government of India has separately developed the Guidelines for the handling of COVID-19 biomedical waste, which needs to be followed besides BiomedicalRules, 2016 so that Corona spread through this can be controlled. Owing to its novel origin and least information about its behaviour, thus it is extremely important to take all precautions possible till we get some medical treatment.

Large language models (LLMs) incur prohibitive computational costs when deployed as monolithic systems for multi-domain query processing. This paper proposes Accelerated LLM, a modular architecture that replaces a single general-purpose LLM with an ensemble of task-specialised small language models (SLMs) governed by a neural query router and a Mamdani fuzzy inference system. The router embeds each user query using a frozen sentence encoder and classifies it across four task domains—summarisation, translation, question answering, and text generation—routing confident queries directly to the corresponding SLM. Ambiguous queries are escalated to a three-input fuzzy logic system operating on Query Length, inter-Domain Overlap Score, and Classifier Confidence, enabling principled handling of imprecise inputs. A reinforcement-learning feedback loop, validated through a controlled pilot deployment, continuously refines the routing policy. The complete pipeline, including the sentence encoder, totals approximately 2.14 billion parameters—a 98.8% reduction relative to GPT-3.5 (175 B). The integration of fuzzy logic into the routing stage raises classification accuracy from 91.5% to 94.3% and reduces the hallucination rate to 9.8% (minor) and 6.4% (major). Evaluated on healthcare-augmented benchmarks against ChatGPT-3.5, Claude, Mistral 70B, and two contemporary compact models (GPT-4o-mini and Llama 3.1-8B-Instruct), Accelerated LLM achieves competitive or superior task-specific performance at a fraction of the parameter count. A small-scale pilot evaluation in the legal domain indicates that the routing and fuzzy logic components retain partial effectiveness beyond the primary healthcare setting, though full multi-domain validation remains future work.