Roughly a year after the OpenShoe implementation was published and after a hectic year of further development, we have now developed a complete real-time multi-agent cooperative localization system based on the setup illustrated in the image above. An article describing the main parts of this system is available at arXiv (Cooperative localization by dual foot-mounted inertial sensors and inter-agent ranging)

This work on cooperative localization is not really a part of the OpenShoe project but it naturally have many connections since the OpenShoe units are an crucial component in it. From the OpenShoe perspective, the main result of the article is that it describes how the OpenShoe units should be integrated/viewed from a system perspective. Also, building the system has forced us update the software of the system and to implement many new features. I hope that I will soon find find the time to clean up these features and merge the banches with the trunk on SourceForge.

• Fus­ing the Infor­ma­tion from Two Nav­i­ga­tion Sys­tems Using an Upper Bound on Their Max­i­mum Spa­tial Sep­a­ra­tion
• Smooth­ing for ZUPT-aided INSs

have now been added under the publication section.

The workshop concerned multiple areas within the localization research conducted at KTH and IISc and therefore naturaly also the OpenShoe project, or rather primarily the use of the OpenShoe units. The people at IISc has recently published the paper Data Fusion of Dual Foot-Mounted INS to Reduce the Sys­tem­atic Head­ing Drift (See Publications) and we had a small benchmark test comparing the solution to the sensor fusion solutions earlier developed at KTH. Benchmark tests going on:

We have been considering buying a 3D printer for quite some time now so I guess we are now running short of excuses.