Wszystkie przebiegi Mapowanie domu GPT-4o mapa 003 Laundry Archiwum

omp-door-exit-luna6-001

dom · homebot-dom · katalog homebot-dom/runs/omp-door-exit-luna6-001

omp-door-exit-luna6-001

dom · homebot-dom · 2026-11-03 09:07 · homebot-dom/runs/omp-door-exit-luna6-001
0/3 w koszu

MP4 · 106.9 KB · pełny ekran w playerze

obiektdostarczonyw koszu
sock_bluenie0%
shirt_rednie0%
towel_greennie0%
skilldoor-exit-ground
modelopenai/gpt-6-luna
profilgpu
LLM / koszt5 calli · $0.0025
ag. limity5 req · 1 seq · 5 tools
czas misji20.3s
Logi
calle LLM
5
tool calls
5
koszt zadania
$0.0025
tokeny (sum)
54,496
in / out
15 / 781
cache R/W
39,729 / 13,971
czas modelu
15.2s
Calle LLM na zadanie
#czasco zrobił (tool + wynik)tokenykoszt
11.5s
toolUse
robot_capabilities
DISCOVER: retrieve current robot capabilities and calibration
mode=SIMULATION_ONLY
8,066
in 3 · out 28 · cR 0 · cW 8035
$0.0010
22.6s
toolUse
mission_memory_read
Read configured mission memory for context; historical notes are unverified
{"revision":0,"agent_notes":null}
robot_observe
SURVEY: observe current RGB-D, visible door/handle masks, sensors and candidates
{"observation_id":"obs-3eba713c-1-350","session_id":"3eba713c","time_s":0.7,"rgb_path":"/home/pawel/Documents/HomeBot_P…
8,888
in 3 · out 80 · cR 8035 · cW 770
$0.0002
32.1s
toolUse
robot_look
Survey visible door and handle; aim camera toward candidate for direct visual confirmation
ok
11,455
in 3 · out 69 · cR 8805 · cW 2578
$0.0004
41.7s
toolUse
robot_observe
SURVEY: inspect updated RGB-D door and handle from raised camera gaze
{"observation_id":"obs-3eba713c-2-1163","session_id":"3eba713c","time_s":2.326,"rgb_path":"/home/pawel/Documents/HomeBo…
11,542
in 3 · out 33 · cR 11383 · cW 123
$0.0001
57.2s
stop
**Status: blocked** — nie potwierdzono otwarcia ani przekroczenia progu. | Etap | observation_id | Wynik ostatniej komendy | Kontrola geometryczna | Następne działanie | |---|---|---|---|---| | DISCOVER | — | Możliwośc…
14,545
in 3 · out 571 · cR 11506 · cW 2465
$0.0007

Źródła: messages.json · openrouter-costs.json. Koszt = OpenRouter total_cost (nie zerowe metadata SDK).

evaluation.json
{
  "objects": {
    "sock_blue": {
      "fraction_in_bin": 0.0,
      "delivered": false,
      "max_z_m": 0.003981342801044341
    },
    "shirt_red": {
      "fraction_in_bin": 0.0,
      "delivered": false,
      "max_z_m": 0.003976715024182752
    },
    "towel_green": {
      "fraction_in_bin": 0.0,
      "delivered": false,
      "max_z_m": 0.003976084453050983
    }
  },
  "events": [],
  "sim_time_s": 2.3259999999999654,
  "physics": "MuJoCo flex 2D FEM/contact/self collision",
  "controller_kind": "see host run manifest; evaluator is not a planner sensor",
  "doors": {
    "g_entrance": {
      "hinge_angle_rad": 7.852686214675603e-38,
      "handle_angle_rad": -4.467852909247484e-18,
      "latch_active": true
    },
    "g_garage_utility": {
      "hinge_angle_rad": -8.548120597307296e-09,
      "handle_angle_rad": 9.868709782741706e-19,
      "latch_active": true
    },
    "g_utility_hall": {
      "hinge_angle_rad": 0.0,
      "handle_angle_rad": -4.590678439863688e-18,
      "latch_active": true
    },
    "g_office": {
      "hinge_angle_rad": -8.548120597294193e-09,
      "handle_angle_rad": 7.800437227746091e-19,
      "latch_active": true
    },
    "g_wc": {
      "hinge_angle_rad": -8.539104001282694e-09,
      "handle_angle_rad": -1.7218714501310728e-17,
      "latch_active": true
    },
    "g_vestibule": {
      "hinge_angle_rad": -8.539104001282694e-09,
      "handle_angle_rad": -1.7218714501310728e-17,
      "latch_active": true
    },
    "g_fitness": {
      "hinge_angle_rad": -8.548120597297659e-09,
      "handle_angle_rad": -6.87372542231557e-17,
      "latch_active": true
    }
  },
  "floor_id": "ground"
}