BMC to now use AI for error-free building proposal submissions in Mumbai

Mumbai: Chief Minister Devendra Fadnavis on Friday launched the AI Building Proposal Management System for the BMC at the Mumbai Tech Week 2026 inaugural session.

BMC commissioner Ashwini Bhide said the new system, called CIVIT TWIN, will help architects submit error-free proposals for approvals, reducing delay and thereby, cost escalation. The impact of this intervention would be a 60% reduction in application rejections, she said, after a presentation on how CIVIT TWIN works.

“It simulates the full BMC scrutiny outflow before the application is submitted by introducing 12 specialised AI agents at different levels of the submission process,” Bhide said. The AI will read all documents submitted, suggest what documents are missing, what is the correct entry, and how the missing documents are to be rectified and submitted. It will also read all drawings and guide the applicant on what kind of law deviation has happened. It will also read city survey maps, GIS maps, how to factor in zoning restrictions and also historical data, said Bhide.

Pointing to the paucity of land in the city, Bhide said all construction routes are through redevelopment.

The municipal commissioner further said that every year, the BMC receives over 8,000 building proposals for permissions and out of these, 40 lakh sq m is approved. “There are 27 additional FSI (floor space index) regulations layered over the base regulation, overlapping spatial zones, not a single plot is easy for redevelopment due to proximity to railways, metros, air funnels, coastal regulatory zones,” said Bhide, while explaining the need for the AI Building Proposal Management System.

A proposal, she said, goes through 25 different departments for NOCs, 60 documents per project require vetting and validation. “Not a single project is possible without concessions,” Bhide added.

Once the proposal is on Auto-DCR, the process is transparent; it can be tracked on the dashboard. The problem still lies in the very process of application. After an application is submitted the architect is not sure whether the plan will be approved as is, which creates delays and cost escalation. An application is submitted multiple times because of the complexity of regulations, said Bhide.

The AI system will make it easy for submission of proposals as it points out errors in plans and helps correct these errors. It makes the automated system predictable.