Invisible Maps

Small-mobility

Once autonomous driving on streets is realized, the next step is to enable autonomous driving by small robots in towns and shopping malls. While navigating in such environments, the robot’s sensors are expected to be blocked by pedestrians. Under such circunstances, sensors that rely on observing geometric structures become inefficient. For such scenarios we prose the use of invisible information.

Invisible Information

In this research we refer to invisible information as those that are not detected by sensors that measure geometric structures (LIDAR, vision, etc.), and includes environmental magnetic fields and radio signal strength of Wifi routers.

Source code

Libraries and example code to process and realize localization using invisible data is available in our github page

Datasets

As part of this research, we have collected several datasets that include 3D point clouds, odometry, Magnetic and Wifi data. These datasets are available here

Funding

• National Institute of Information and Communications Technology
  

  研究開発課題名   データ連携・利活用による地域課題解決のための実証型研究開発（第2回） 
        副題           スモールモビリティによるラストワンマイル達成のための混雑環境でもロバストな不可視地図のオープン化 
        研究開発期間    2019年9月24日～2021年3月31日 
        受託者名        国立大学法人宇都宮大学<代表研究者> 
          

Related Publications