[evaal contest] more clarifications

Dario Salvi dsalvi at lst.tfo.upm.es
Mon Jun 25 15:54:11 CEST 2012

Dear Julia,

first of all I would suggest to send questions to the mailing list so 
that all competitors are aware of our suggestions. For this reason I am 
copying this reply to the contest list.

>     Basically, we should be fine doing both measurements and placement
>     of the sensors within 1 hour. The only thing I would really like to
>     check in advance is that our velcro can be easily peeled off of the
>     ceiling tiles. In order to do that I don't really need to enter the
>     lab, all I need is a single tile made of the same material used for
>     the ceiling. Maybe you have a sample of those tiles that we could
>     try it with, or otherwise perhaps someone could just take down one
>     of the tiles in the lab and take it outside where we would apply one
>     strip of velcro and see whether it peels off easily.
>     We are arriving July 2 and we are scheduled on July 5 so that if we
>     had a possibility to try it before July 4, we would have some time
>     for a backup plan in case we find out that our velcro is damaging
>     the surface. Even though our team is scheduled the last in the
>     localization track, we really care not to leave any damage after us!

I think it is fair to show you the tile before the competition so that 
you can find the best solution for it that would not damage the tile itself.

You can come before your turn, choose the day you prefer, and I will 
show you a tile (if you want you can bring one to make tests somewhere). 
Just tell me when precisely you are going to come and remember that you 
will be not allowed to actually enter the lab (we will meet outside).

>     It's made clear in the refinement document that we should bring all
>     the tools for bricolage with us, and we are definitely going to do
>     so. One thing that we cannot bring is a ladder, and I was assuming
>     that you will have it in the lab, is this correct?

I have just sent an email to the list regarding this point with a list 
of tools we are going to provide, the ladder is part of it.
If you have any other requirement, please tell us before Wednesday.

>     I also have a question regarding phase 3 where we have to localize
>     one person in the presence of the other one. Our initial plan was to
>     have the second person wear an RFID tag, however, we had technical
>     difficulties with the RFID development track and had to abandon this
>     plan. Our main system based on only motion sensors placed with
>     specific topology in mind allows for recognition of two moving
>     objects so that we would be able to track and report locations of
>     both of them. The only issue here is that we don't have a means to
>     distinguish between the person who has to be tracked and the other
>     "noise" person. Will any information be given about their initial
>     positions or time when they start moving?
>     I am aware that you cannot disclose the paths of the actors before
>     the competition, so what I am asking is whether it's possible to
>     know some coarse estimate of where both actors start their paths
>     (e.g., the trackable person starts moving anywhere inside the lab,
>     the "noise" person starts from the porch and then enters the lab) or
>     whether there is any time gap between two actors' initial movements
>     (e.g., the trackable person starts moving 5 seconds earlier than the
>     "noise" person). If none of these is available to our system, we
>     will have to blindly pick one of the paths to report and therefore
>     will have only 50% chance to guess right. We could also report both
>     actors' paths if this was acceptable.
>     Note, in the real world scenario, if our system was to track the
>     patient inside the apartment and then the visitor (e.g. a caregiver)
>     entered, we would know the initial position of the patient from the
>     localization information available to the system before the visitor
>     enters. Essentially, our system would know both: coarse estimate of
>     both persons' starting positions and a time gap between the moments
>     when they first appeared on the localizable map of the apartment.

We have discussed about this point, Francesco Potortí will reply to this 
point in this mailing list very soon.

If you have any other question don't hesitate in contacting me.


>     Thank you in advance,
>     Julia Vlasenko

Dario Salvi
+34 915495700
dsalvi at lst.tfo.upm.es
Life Supporting Technologies
Departamento de Tecnología Fotónica y Bioingeniería
Universidad Politécnica de Madrid
Madrid, Spain

More information about the Contest mailing list