OpenShoe

As I have pre­vi­ously noted in a post, we have recently been devel­op­ing a massive-MIMU (multi-IMU) plat­form from single-chip IMUs. (The abbre­vi­a­tion is par­tially a play with the com­mon MIMO (multiple-input-multiple-output) abbre­vi­a­tion within the field of com­mu­ni­ca­tion and also the term massive-MIMO to describe MIMO sys­tems with alot of anten­nas.) At the move­ment the plat­form looks like this:

The plat­form will be pre­sented on the upcom­ing ISISS con­fer­ence 25–26 Feb 2014 in Laguna Beach CA, USA. Massive-MIMU is to some extent a sep­a­rate research track but also con­sti­tute an exper­i­men­tal plat­form for the devel­op­ment of new Open­Shoe modules.

The plan is to release the whole plat­form open-source under the Massive-MIMU sec­tion (and of cause on source­forge). The soft­ware part is already in place at source­forge, specif­i­cally in this file. This we have been run­ning and devel­op­ing since the autumn. We are cur­rently work­ing on a sec­ond gen­er­a­tion of the MIMU plat­form which will hold 32-IMUs (which will truly make it a massive-MIMU plat­form). How­ever, through­out the devel­op­ment we have come to real­ize that we want to make some changes which will affect the soft­ware. Con­se­quently, we have been a lit­tle reluc­tant to put up the old gen­er­a­tion of boards since we do not place to sup­port them in the future. If you really want the schematic and lay­out of the old boards, please send us an e-mail and I will be happy to pro­vide it. Oth­er­wise, I hope you can be patient until we fin­ish the new generation.

Last week, I finally fin­ished and pushed some updates/new devel­op­ment of the soft­ware onto Source­Forge. What has been con­structed is what I call an iner­tial fron­tend. The basic idea is:

In words it’s a pro­cess­ing block (group of rou­tines together with some state vari­ables) which pack­age and per­forms all the pro­cess­ing prior to the iner­tial nav­i­ga­tion, i.e. cal­i­bra­tion com­pen­sa­tion (and fusion of array data), test sta­tis­tics cal­cu­la­tion and thresh­old­ing for the ZUPTs, and some online bias esti­ma­tion. This is pro­cess­ing which is tightly con­nected with the IMU and rather inde­pen­dent of the iner­tial nav­i­ga­tion. The most sig­nif­i­cant attrib­utes of the fron­tend is that it’s com­pletely imple­mented with inte­gers arith­metic (so that it could poten­tially be placed in a pro­cess­ing unit embed­ded with the IMU) and the test sta­tis­tics is imple­mented in a recur­sive form (which makes the com­pu­ta­tion sig­nif­i­cantly less expen­sive to start with and also inde­pen­dent of the win­dow size). This far it seems to be work­ing great. We’ve used a crude ver­sion of it since april last year so I don’t expect any larger prob­lems with it but be aware that it’s not been thor­oughly tested yet. This far the code and the under­ly­ing the­ory is pretty much only doc­u­mented in var­i­ous tech­ni­cal notes on my com­puter but hope­fully I will have time to gather it in a paper soon. Any­way, now you know roughly what the 400-line block of obscure code is doing.

We are cur­rently in a period of inten­sive devel­op­ment. We have now used our units for over two years. Dur­ing this period we have learnt a lot and the devel­op­ment on the sen­sor mar­ket has been sig­nif­i­cant. To take advan­tage of this we are now devel­op­ing a new gen­er­a­tion of the mod­ules as well as mak­ing sig­nif­i­cant updates to the embed­ded soft­ware. The tar­geted fea­tures for the new mod­ules are:
– Sin­gle chip IMUs (lower cost and pro­file)
– Wire­less and bat­tery pow­er­ing
– Inclu­sion of a mem­ory
– Mag­ne­tome­ters
After a stopover at IISc, John-Olof is at IIT Kan­pur (India) since a week back in time. Here the tar­get is to carry out the devel­op­ment in coop­er­a­tion with GT Sil­i­con Pvt Ltd, an elec­tronic sys­tems design com­pany with whom we have had a long relation.

Cur­rently, we are in a PCB design and fab­ri­ca­tion phase. The IMU array boards which were briefly men­tioned in this post is to some degree a pre­de­ces­sor of these mod­ules (as well as a sep­a­rate track in terms of research) and have been act­ing as test­boards for some embed­ded software-hardware design ideas. In par­al­lel with the devel­op­ment with the new mod­ules, we are work­ing on the soft­ware. Some of the soft­ware updates are already avail­able on Source Forge. For exam­ple sup­port for the IMU array is already there. Cur­rently a com­plete mea­sure­ment fron­tend is under development.

More infor­ma­tion will be posted shortly.

As pre­vi­ously said, we have used the Open­Shoe units to build a com­plete infra­struc­ture free coop­er­a­tive local­iza­tion sys­tem. A cou­ple of month ago we did some lit­tle larger field tests for the first time, with smoke­divers from a nearby firede­part­ment. One of the tests runs can be seen in the fol­low­ing vide clip. This gives a high level exam­ple of the poten­tial of the imple­men­ta­tion. Enjoy!

At the last IPIN con­fer­ence in France we pre­sented two papers:

• Voice radio com­mu­ni­ca­tion, pedes­trian local­iza­tion, and the tac­ti­cal use of 3D audio
• Stand­ing still with iner­tial navigation

These have now been added under the pub­li­ca­tions section.

The for­mer paper con­cerns how the com­plete coop­er­a­tive local­iza­tion sys­tem and head-related-transfer func­tions (HRTFs) may be exploited to pro­vide a spa­tial per­cep­tion of rel­a­tive loca­tions within a group. Sim­ply, an agent will here where he has his col­leagues when they talk and vice versa. The sec­ond arti­cle con­cerns a small detail in the mech­a­niza­tion imple­men­ta­tion which will pre­vent the foot-mounted iner­tial nav­i­ga­tion sys­tem from drift­ing when an agent is stand­ing still.

The arti­cle Coop­er­a­tive local­iza­tion by dual foot-mounted iner­tial sen­sors and inter-agent rang­ing has now gone into print and can be found open access on EURASIP’s site. (Pre­vi­ously it was only avail­able from arXiv.) The same goes for the Recur­sive Bayesian ini­tial­iza­tion of local­iza­tion based on rang­ing and dead reck­on­ing paper.