Niti Aayog mulls use of AI to manage crowd during Kumbh
At a time when stampedes during religious congregations have taken hundreds of lives in India, be it the incident in Datia in Madhya Pradesh in October 2013 when 115 persons were killed as a stampede had broken out on a bridge near a temple during Navaratri, or in January 2011 when 106 devotees were crushed to death in Sabarimala in Kerala, Niti Aayog has strongly propagated use of artificial intelligence ( AI) as an effective solution for crowd management, especially in events like “Kumbh Mela”.
With the “ardh Kumbh Mela” scheduled to begin in January 2019 in Allahabad, the Narendra Modi- led dispensation is quite keen to use AI as a tool to prevent any such man made tragedy during one of the world’s largest religious gathering.
AI is a concept which is aggressively being pushed by the NDA dispensation and the government think tank has been asked to prepare a blueprint on its wider usage in various fields. Under this endeavour, Niti Aayog has suggested that it can be quite useful for crowd management, especially in a country which has witnessed several tragedies occurring due to poor crowd control tactics by agencies.
Use of AI in providing effective solutions in crowd management in recent times have been in vogue globally, and has given fruitful results in averting city- scale challenges such as managing mega footfall events, emergency and disasters, the Aayog has suggested in its proposal.
Giving example of how Accenture worked with the Singapore government during their 50th anniversary of independence and developed a solution aimed at predicting crowd behaviour and potential responses to incidents, Niti Aayog has said that it resulted in achieving 85 per cent accuracy in high crowd activity, crowd size estimation and object detection.
This could be replicated in India in religious congregations like during Kumbh Mela when AI and big data solutions can help in predicting crowd behaviour and avert possibility of a stampede. Similar Big Data and AI solutions could help with advance prediction and response management during such mass level gatherings.