Help me funding my research about autonomous mapping

A project by: Cristian Hernandez

£420
pledged of £10,153 target

This project did not reach its target.

Completion Date: Tue 08 Sep 2015
4 years ago

Now you can read my Research Proposal. already accepted for the MRes in Spatial Data Science and Visualization in UC, here:


Understanding and Visualizing Data from Autonomous Agents in Urban Environments

I would like to share my research proposal for the MRes Spatial Data Science and Visualization in University College London. -

The emergence of “Intelligent Agents” (1), in terms that artificial intelligent artifacts are acquiring negotiation capacity, represent a new challenge to scientist studying urban environments under the Smart City discussion.  

In this sense, concepts from different companies and research centres around the world have gave us a view and taste of how Autonomous Systems (cars, robots and other objects - 2) learn from the space and create maps in real time, using geolocated data harvested in the last few years, several sensors and SLAM’s (Simultaneous Localization and Mapping) algorithms. 

The research topic here presented aims to understand how Autonomous Agents are learning from the space (from the geolocated database to sensors), how we are creating artificial intelligence to map our cities (SLAM algorithms), and which data these agents are collecting (which variables, zones into the city, among others).    

The importance of study Autonomous Agents in urban environments lies in the growing adoption in cities of intelligent technologies to resolve arising problems as mobility, for example. Thus, to research how the “autonomous mapping process” will allow us to read our cities in a different way, presenting new challenges to urban scientists, planners and designers.  

To achieve the research topic, this proposal is based on the experimentation with different low cost technologist (3) as Raspberry Pi, Arduino or Google Tango devices, which are designed to incorporate different sensors, and are designed to be programmable using SLAM algorithms, through languages like Python, C (#,++) or Java, in specific libraries as OpenSlam (4), who contain a collection of SLAM algorithms, some of them open source, as “Gmapping”. 

Some possible outcomes from this research line are, for example: 

  • Improve SLAM algorithms or create new ones.
  • Improvement or creation of new sensors to feed SLAM algorithms, aims to process and create new maps. 
  • Create new autonomous agents with “mapping capacity”, as bicycles, buses, houses, buildings, traffic signs, etc. 
  • Improve the city itself for autonomous agents, in terms of connect on a network different objects in the city, or allow urban artifacts to talk with autonomous agents, as cars for example. - Use SLAMs as a new way to process spatial data from sensors already installed in the city.   

Bibliography: 

1.- Wooldridge & Jennins, “Intelligent Agents”. 1995
2.- Grisetti et al, “A Tutorial on Graph-Based SLAM”. 2010
3.- Mano:Chang:Wengrowski, “Remotely-Processed Visual SLAM Using Open-Source Software”, 2013.
4.- See https://www.openslam.org/  
5.- Frischenschlager, “Introduction to Simultaneous Localization and Mapping”, 2013



4 years ago

Video de la campaña en Español:


https://vimeo.com/135880668

4 years ago

Here you can see my Pitch on video... I don't know why is not published on my actual pitch. Anyway:

Watch the video HERE.