## playerstage-users

 [Playerstage-users] Uncertainties in measurement From: varunraj - 2007-10-25 19:21:37 ```hi, I coded a deterministic multi-robot algorithm which makes the robots to meet at a point. i found that for each run there is a difference in measurement of the final rendezvous point with an error of +/- 0.3-0.7 units. I tried analyzing this behavior by just using one robot and finding out the distance it moves with each iteration. Well, sometimes it moves by one step or sometimes it doesn't. So, is this uncertainty built into the player/stage to emulate the real-world? when i ran the same algorithm in matlab using a point mass model, there was no error. with every run they converged to the same point. thank you Varun Raj -- View this message in context: http://www.nabble.com/Uncertainties-in-measurement-tf4693018.html#a13413986 Sent from the playerstage-users mailing list archive at Nabble.com. ```
 Re: [Playerstage-users] Uncertainties in measurement From: Brian Gerkey - 2007-11-01 22:09:48 ```On Oct 25, 2007, at 12:21 PM, varunraj wrote: > I coded a deterministic multi-robot algorithm which makes the > robots to meet > at a point. i found that for each run there is a difference in > measurement > of the final rendezvous point with an error of +/- 0.3-0.7 units. I > tried > analyzing this behavior by just using one robot and finding out the > distance > it moves with each iteration. Well, sometimes it moves by one step or > sometimes it doesn't. So, is this uncertainty built into the player/ > stage to > emulate the real-world? when i ran the same algorithm in matlab > using a > point mass model, there was no error. with every run they converged > to the > same point. The differences are probably due to the fact that your control program is decoupled from, and asynchronous with respect to, the simulation. Small differences in timing (e.g., when you send a command, when you process a sensor reading) can produce small differences in behavior. This is not a feature that we designed in, but rather an artifact of the architecture, in which the control program and the robot (whether real or simulated) are decoupled. This is how things happen with real hardware; in your control program, you get to asynchronously sample some physical process (e.g., a laser that is taking scans, a microcontroller that is integrating odometry) that operates at own frequency. If you're looking for a perfectly repeatable simulation, then I recommend using Stage without Player (via the libstage API). brian. ```
 Re: [Playerstage-users] Uncertainties in measurement From: varunraj - 2007-11-04 06:21:20 ```Hi, well, its kinda nice that it happens, gives a feeling of a real robot implementation. i expected it but wanted to be sure before i write it up in my report. thank you... varun Brian Gerkey wrote: > > > On Oct 25, 2007, at 12:21 PM, varunraj wrote: > >> I coded a deterministic multi-robot algorithm which makes the >> robots to meet >> at a point. i found that for each run there is a difference in >> measurement >> of the final rendezvous point with an error of +/- 0.3-0.7 units. I >> tried >> analyzing this behavior by just using one robot and finding out the >> distance >> it moves with each iteration. Well, sometimes it moves by one step or >> sometimes it doesn't. So, is this uncertainty built into the player/ >> stage to >> emulate the real-world? when i ran the same algorithm in matlab >> using a >> point mass model, there was no error. with every run they converged >> to the >> same point. > > The differences are probably due to the fact that your control > program is decoupled from, and asynchronous with respect to, the > simulation. Small differences in timing (e.g., when you send a > command, when you process a sensor reading) can produce small > differences in behavior. > > This is not a feature that we designed in, but rather an artifact of > the architecture, in which the control program and the robot (whether > real or simulated) are decoupled. This is how things happen with > real hardware; in your control program, you get to asynchronously > sample some physical process (e.g., a laser that is taking scans, a > microcontroller that is integrating odometry) that operates at own > frequency. > > If you're looking for a perfectly repeatable simulation, then I > recommend using Stage without Player (via the libstage API). > > brian. > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Splunk Inc. > Still grepping through log files to find problems? Stop. > Now Search log events and configuration files using AJAX and a browser. > Download your FREE copy of Splunk now >> http://get.splunk.com/ > _______________________________________________ > Playerstage-users mailing list > Playerstage-users@... > https://lists.sourceforge.net/lists/listinfo/playerstage-users > > -- View this message in context: http://www.nabble.com/Uncertainties-in-measurement-tf4693018.html#a13570344 Sent from the playerstage-users mailing list archive at Nabble.com. ```