Fire Safety in Indian Coal Mines using Machine Learning Techniques
Saaniya Qaiser1, Anurag Sharma2, Harini Murugan3

1Saaniya Qaiser, Student, Computer Science and Engineering from SRM Institute of Science and Technology, Kattankulathur, India.
2Anurag Sharma, Student, Computer Science and Engineering from SRM Institute of Science and Technology, Kattankulathur, India.
3Mrs. Harini Murugan,  Assistant Professor, Information Technology, SRM Institute of Science and Technology, Kattankulathur, India.
Manuscript received on January 22, 2020. | Revised Manuscript received on February 05, 2020. | Manuscript published on February 29, 2020. | PP: 4003-4005 | Volume-9 Issue-3, February 2020. | Retrieval Number:  C6418029320/2020©BEIESP | DOI: 10.35940/ijeat.C6418.029320
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (

Abstract: There are around 493 coal mines in India (300+ underground and around190 opencast mines) engaged in coal production for meeting energy and other requirements of our country. Coal and the process of mining itself creates an environment conducive for self-oxidation leading to build up of heat and subsequently break out of fire. This causes safety hazards, decrease in production, increased in de-settlement of colonies, fire related fatalities and risk to life and property. Occurrence of fires in coal mines has always been an undesirable proposition for the coal mining community worldwide due to its high hazard potential towards loss of human lives and property. However, with advent of AI/ML and deep learning, there emerges a vast scope of leveraging its application towards significantly reducing fire hazards in coal mining. Data capturing from such fiery mines, providing machine learning and predicting it beforehand for similar mining situations would significantly enhance safety standard in coal mining industry. This project proposes to develop an algorithm on getting input data from the past incidences/accidents of fire in coal mines and apply machine learning software to help it learn pattern/features vis a vis the fire outcomes. Once the learning is over and data trained, the programme would process the test data of other active projects and may predict for fire threat during forthcoming mining operation. The algorithm aims to enable mining personnel to assess and evaluate the risk of fire in their workplace and take informed decisions based on the predictions based on Machine learning outputs. Also, active fires can as well be studied and predicted in a similar way. This will help the mining team to decide about the right approach of continuing mining operation in such an affected area.
Keywords: Open cast ventilation pillar, Logistic regression, spontaneous combustion.