AfterCnesTODOs » History » Revision 71
Revision 70 (Pierre Narvor, 2018-10-16 12:09) → Revision 71/72 (Simon Lacroix, 2018-10-25 15:17)
h1. AfterCnesTODOs Big TODO list following the CNES tests, sorted in topics, for which a responsible and a co-responsible is named. * Hardware (Simon, Pierre) ** Minnie *** -Redresser les HazCam- *** -Install minnie-superbase- *** -Install rear Hazcams- *** -Tout revoir le câblage- *** -Get some working Ethernet cables- *** Surveiller l'évolution de la température des deux CPUs -boucher l'ouverture arrière du boitier pour une meilleure ventilation- *** Câbler les synchros pour toutes les caméras *** Trouver un moyen démontable de surélever l'antenne Wifi *** Fix the GPS issue: where should we put the antenna? *** Ré-installer le Lidar Sick *** Acheter des cameras avec auto-exposure ? *** -Réparer le câble de la MTI- *** Install the dual antenna GPS ? (where ? with a deployable mechanism ?) *** Documenter le câblage définitif *** Tests extensifs Joystick ** Mana *** -Installer la carte USB 3.0- *** -Installer la caméra HS-D- *** Changer le disque dur ? *** Trouver un moyen démontable de surélever l'antenne Wifi *** Tests extensifs Joystick ** Others misc things on the robots hardware *** -5 replacement batteries for the Segway bases- (received) *** Test the wifi backbone and the second access point *** Revoir le montage de la première base GPS *** Faire le montage de la seconde base GPS *** Déployer le seconde base GPS *** Moyen de géoréférencer un point (fil à plomb - mètre, clou de géomètre) *** Test the SBG sensor ? *** -Faire le bilan de toutes les tourelles PTU- *** Get a few spare cables with connectors of the robots alimentation board (some with banana plugs) * Logistique (Simon, who ?) ** Things to buy *** New Wifi setup (order placed) *** Vacuum cleaner *** Really easily deployable tent (like the one Onera had) *** Solid "sardines" and a strong hammer *** A more solid light camping table? *** Two higher masts for the two Wifi access points *** Electric current generator *** Small gasoline tank *** Backup harddrive for the on-board PCs (are they all of same type?) *** Hard drive with all the backups of Minnie, Mana, robotpkg, Ubuntu... et tout et tout (un RAID ? Un graveur de DVD ? Des clefs USB, autres moyens ?) *** A second hard drive (more than one?) to store the acquired data ** Trier et ranger la boite à outils ** Tester le réseau Wifi longue portée (avec 2nd access point) ** Make the definite list of all that we need to bring *** Matériel de backup: bastide, access points, PTU, Joysticks, GPS, minnie-base... *** Matériel pour le calibrage extrinsèque (barres RexRoth?) * Déploiement sur le terrain (Ellon, who ?) ** Update Bastide software ** -Declare minnie-cpu on bastide- * Aspects système sur les robots (Ellon, Quentin) ** Faire en sorte que les devices USB soient toujours sur le même port (avec udev something) ** Faire des miroirs d'Ubuntu ** Mount all the minnie-superbase hard drive on minnie-cpu ** Synchronise both robots time * Scripts de startup sur les robots (Ellon, ...) ** -Define the way to do them- ** Finish them * Project deliverables (Simon, ...) ** 20/08/2018: 24 deliverables are due by the end of the project !?! * RobotPkg (Ellon, Quentin) - mais tous doivent savoir faire ** -Packager GPS- ** -Packager velodyne-genom3- ** -Pacakger infuse-asn1-conversions- ** Packager PTU ** -Packager Envire- ** -Packager InFuse types- ** -Package infuse-pom-base- ** -Packager infuse-pom-crisp- ** -Packager infuse-crisp-genom3- ** -Packager infuse-pom-tokamak- ** -Packager infuse-tokamak-ros- ** -Packager infuse dem building- ** -Packager infuse dem building ROS- * ROS ** Possibilité (intérêt) de séparer des ros-core ? ** Utilisation de nodelets ? * CDFF (Quentin, who ?) ** -Finalize the InFuse-types- ***- Introduce flags in the pose type to specify which variables are estimated- *** -introduce a string in the pose type to specify who produced the pose- *** -There was an issue in the serialization of strings?- *** -Get rid of this @#$$#% dependency on a .exe file !- ** -Move the types to the main CDFF -- clarifies the situation after the early August merge wrt our package.- ** DFNs (& DFPC ?) ** -Make separate git repos- * Intégration / exploitation des softs de MAG (Andrea, who) ** Récupérer un docker à jour * Display utilities => *Visualization is currently being developped by DFKI and necessitates subscription to DFNCI/DFPCCI* ** Integrate the DFKI developments? ** Where would RVIZ be helpful? ** Develop a web-based monitoring system? * Gestion des jeux de données acquises (Simon, who) ** Dump rosBags, or have the modules dump their data individually? => *This needs to be discussed seriously* ** Définir la manière de gérer les données sauvegardées ** Faire des utilitaires d'exportation des rosBags vers des fichiers individuels ** Le service replay de DFKI est-il OK? * Simulation (Andrea, ...) ** Make a clear scheme of the interfaces (the ROS graph, and a more conceptual one?) Will be done by MAG * Les modules/noeuds/DFPC ** Cameras (MAG) *** Have a node able to acquire any of the three stereo benches, including "all of them"? *** Discuss with MAG the best trigger choice (controlled, or periodic?) *** Make the extrinsic calibration ** GPS (Ellon, ...) *** Procédure d'initialisation de l'angle wrt. nord *** -Récupération des trames bestutmb- *** Decode the standard deviations *** (Aug. 6) Why can't the bestutmb messages be retrieved at 20Hz? *** Maybe Switch to OEM6 models ? *** Clarify the situation wrt. gps_send / dgpssend, gps_receive... (the later being installed via the Genom2 GPS module!) *** Why the z value are not consistent between Mana and Minnie ? ** RMP's (Pierre, Ellon) *** Comment remonter le fait que la puissance n'est pas en route ? *** -Erreurs du gyro quand le module est en track-mode ?- *** -Procédure d'estimation du biais du gyro- *** Procédure d'estimation des biais de la MTi pour les angles d'attitude *** -RMP4?0 en Odo 3D (récupération des données des MTI)- *** Remonter l'état de la batterie (RMP440 sur Minnie), faire une alarme ? *** Régler la dynamique des Joysticks, limiter à 1.0 m/s à la main *** Tester les fonctions de l'arrêt d'urgence (et notamment re-coder la reprise avec le bouton 3 ?) *** Tester le log des données (faire un log qui dumpe toutes les données brutes ?) *** Define a noise model for the gyro and the MTI ? (doing statistics, at rest and while moving) ** Lidar Velodyne (Ellon, who) *** Define blind sectors, remove associated data *** Check why the 1Hz frequency could ne be reached during tests in the parking *** Dépendance de velodyne-libs sur Envire ? *** Make a permanent acquisition task? ** Lidar Sick (Simon, ?) *** Recover the modules developed by Harold *** Export an InFuse point cloud ** Caméra HS (Edouard, Simon) *** Install acquisition drivers on Mana ** PTU (Pierre, ...) *** Package the ROS node used during CNES tests *** -Export an inFuse-pose- *** -Check the maximum possible speed- : it is 0.5 rad/s *** What is the difference between setpoints and measured angles? Get the measured angles ** Crisp (Pierre, ...) *** -Conclude on the use of Envire- *** -Check the chain of frames- *** -Fix time stamps- *** -Calibration: Calibration: estimate the true values of the fixed frames *(CONTACT MAG about rectified calibration data)*- data)* *** Make it pulse at fastest input instead of fixed frequency *** -Crisp Crisp eat up to 50% of a single CPU core of minnie (too much, to investigate)- investigate) *** DFNization to be discussed *** Pose logging ? ** Tokamak (Quentin, ...) *** -Define the simple "aggregate-crush" fusion scheme- *** Create a fusion interface (even for simple fusion) *** Update set_pose to add a childFrameId *** Update initialization to be able to connect odom frames easily *** Integrate a KF (e.g. to fuse Odometry and VO) ** DEM (Quentin, ...) *** Solve the big bug *** -Review the way the InFuse types are used?- *** Update/Debug tiling system *** Store information in raw form *** Allow Replay *** Integrate luminance information *** DFNCI-ize the DFNs *** DFPCCI-ize the DFNCI-ized DFNs *** quid si update a posteriori du cap? --> stocker les PtClouds et non les DEM locaux ? ** VO (Andrea, ...) *** Make it run on-board the robot *** Infuzation ? ** PG-SLAM (Ellon, ...) ** AbsLoc (Andrea, ...) *** RdV avec Andrea pour définition ** P3D (Simon, Pierre) ** PurePursuit *** Extract it from the docker? * Infusing the DFPCs (Pierre, Quentin) ** Update DEM building with DFNCI, DFPCCI, and new ros nodes : inclusion in CDFF, CDFF_ROS ** Update Tokamak with DFNCI, DFPCCI, and new ros nodes : inclusion in CDFF, CDFF_ROS ** Update CRISP building with DFNCI, DFPCCI, and new ros nodes : inclusion in CDFF, CDFF_ROS * Documentation (all, depending on their responsibilities) ** Mana and Minnie wiring ** Robot startup and security (do not forget to look around the robot wether there are any trailing cable) ** Joystick: procédure d'appariement, de calibrage (jscal -c /dev/input/js0) ** Basic operations (security, start-up) ** Firow vs. wiki-robots vs. redmine ? ** Make a clear scheme of all the involved modules (the ROS graph, and a more conceptual one?)