The goal of this project, called VIM (Voice Indoor Maps) is to implement a prototype of Voice Indoor Map service at the UN HQ buildings in NYC. This VIM service provides indoor map information to Visually Impaired People (VIP) with total visual impairment. The indoor map information is given as verbal instructions, which include POIs about nearby landmarks and safety and turn-by-turn instructions for indoor navigation from point A (or the current location) to B in the indoor space.

The service environment is based on smart phones (Android and iOS) and additional mobile devices if necessary. The architecture for this service consists of two parts – VIM service layer and Indoor Positioning Solution (IPS) layer. VIM service layer is separated from IPS because most IPS are proprietary and the performance and accuracy of each IPS depends on site that no one single IPS could work for all cases. Actually, a hybrid IPS is used for this project, it can be replaced with any other IPS. It is not practical to employ one specific IPS for the VIM. While we could assume IPS as an open source or a proprietary solution, the VIM service layer is developed as an open source software and communicates with IPS through a predefined interface protocol. The source code of VIM application layer is available at https://github.com/STEMLab/VIM.

The VIM service receives indoor maps of a specific building or indoor space in IndoorGML, which is an OGC standard for indoor spatial information. Once the starting point and the destination are given by a voice assistant user-interface of smartphone, the VIM computes the routing considering not only the optimal path but also several constraints. While the proof of concepts test is delayed due to COVID-19 pandemic since 2020, we expect it would be resumed in a near future and extended not only to the UN Headquarter but also other UN facilities in the world.