Abstract—We are concerned with enabling truly large scale autonomous navigation in typical human environments. To this end we describe the acquisition and modeling of large urban spaces from data that reflects human sensory input. Over 181GB of image and inertial data are captured using head- mounted stereo cameras. This data is processed into a relative map covering 121 km of Southern England. We point out the numerous challenges we encounter, and highlight in particular the problem of undetected ego-motion, which occurs when the robot finds itself on-or-within a moving frame of refer- ence. In contrast to global-frame representations, we find that the continuous relative representation naturally accommodates moving-reference-frames – without having to identify them first, and without inconsistency. Within a moving-reference-frame, and without drift-less global exteroceptive sensing, motion with respect to the global-frame is effectively unobservable. This underlying truth drives us towards relative topometric solutions like relative bundle adjustment (RBA), which has no problem representing distance and metric Euclidean structure, yet does not suffer inconsistency introduced by the attempt to solve in the global-frame.

This shows the 121-km path taken between Oxford in the upper left and London in the bottom right. We compute visual estimates for 89.4% of this trajectory and fall back on inertial sensing for the remainder.

This shows the 121-km path taken between Oxford in the upper left and London in the bottom right. We compute visual estimates for 89.4% of this trajectory and fall back on inertial sensing for the remainder.

 

13-km path taken around Oxford with inset showing New College portion (a separate 2.2-km trajectory around New College is also shown.  This sequence is ideal for testing loop-closure detection.

13-km path taken around Oxford with inset showing New College portion (a separate 2.2-km trajectory around New College is also shown. This sequence is ideal for testing loop-closure detection.

 

 

Processing data collected from human-like move- ment in urban spaces (right) is very different, and substan- tially more challenging than processing stable robot data (left). In both settings we capture 512x384 grey scale images at 20Hz.

Processing data collected from human-like move- ment in urban spaces (right) is very different, and substan- tially more challenging than processing stable robot data (left). In both settings we capture 512×384 grey scale images at 20Hz.


  • [PDF] G. Sibley, C. Mei, I. Reid, and P. Newman, “Planes, Trains and Automobiles ñ Autonomy for the Modern Robot,” in In Proceedings of the IEEE International Conference on Robotics and Automation, Anchorage, Alaska, USA, 2010, pp. 285-292.
    [Bibtex]

    @inproceedings{Sibley2010,
    Address = {Anchorage, Alaska, USA},
    Author = {Gabe Sibley and Christopher Mei and Ian Reid and Paul Newman},
    Booktitle = {In Proceedings of the IEEE International Conference on Robotics and Automation},
    Keywords = {Relative SLAM},
    Month = {May},
    Note = {05},
    Pages = {285-292},
    Pdf = {http://www.robots.ox.ac.uk/~mobile/Papers/1998.pdf},
    Title = {Planes, Trains and Automobiles {\~n} Autonomy for the Modern Robot},
    Year = {2010}}