feat(mosaico_integration): add fleet comparison cells to notebook

Two new cells at the end of the notebook that pull LaserScan stats from
both LIDAR_SIM sequences and plot noise vs drift side by side.

Author: Michał Fąferek

demos/mosaico_integration/notebooks/mosaico_demo.ipynb

@@ -239,6 +239,11 @@
     "            )"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "source": "## 7. Fleet comparison: noise vs drift\n\nThe fleet variant injects two different LiDAR faults, both reported as `LIDAR_SIM`. Metadata alone cannot tell them apart. Pulling the actual scan statistics from Mosaico reveals two distinct failure signatures.\n\nSkip this section if you only ran the single-robot variant (it needs 2+ LIDAR_SIM sequences).",
+   "metadata": {}
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -247,6 +252,13 @@
    "source": [
     "client.close()"
    ]
+  },
+  {
+   "cell_type": "code",
+   "source": "if len(lidar_sequences) >= 2:\n    def _pull_scan_stats(seq_name):\n        _sh = client.sequence_handler(seq_name)\n        _topic = _sh.get_topic_handler(\"/sensors/scan\")\n        recs = []\n        for item in _topic.get_data_streamer():\n            arr = np.array(item.data.ranges, dtype=float)\n            fin = arr[np.isfinite(arr)]\n            recs.append({\n                \"timestamp_ns\": item.timestamp_ns,\n                \"range_mean\": float(fin.mean()) if fin.size else math.nan,\n                \"range_std\": float(fin.std()) if fin.size else math.nan,\n            })\n        df = pd.DataFrame(recs)\n        df[\"t_s\"] = (df[\"timestamp_ns\"] - df[\"timestamp_ns\"].iloc[0]) / 1e9\n        return df\n\n    seq_a, seq_b = lidar_sequences[0], lidar_sequences[1]\n    df_a = _pull_scan_stats(seq_a)\n    df_b = _pull_scan_stats(seq_b)\n\n    fig, axes = plt.subplots(2, 2, figsize=(14, 6), sharex=\"col\")\n\n    axes[0, 0].plot(df_a[\"t_s\"], df_a[\"range_std\"], color=\"crimson\", lw=1.4)\n    axes[0, 0].set_ylabel(\"range std [m]\")\n    axes[0, 0].set_title(seq_a, fontsize=9)\n    axes[0, 0].grid(alpha=0.3)\n\n    axes[0, 1].plot(df_b[\"t_s\"], df_b[\"range_std\"], color=\"crimson\", lw=1.4)\n    axes[0, 1].set_title(seq_b, fontsize=9)\n    axes[0, 1].grid(alpha=0.3)\n\n    axes[1, 0].plot(df_a[\"t_s\"], df_a[\"range_mean\"], color=\"steelblue\", lw=1.4)\n    axes[1, 0].set_ylabel(\"range mean [m]\")\n    axes[1, 0].set_xlabel(\"time [s]\")\n    axes[1, 0].grid(alpha=0.3)\n\n    axes[1, 1].plot(df_b[\"t_s\"], df_b[\"range_mean\"], color=\"steelblue\", lw=1.4)\n    axes[1, 1].set_xlabel(\"time [s]\")\n    axes[1, 1].grid(alpha=0.3)\n\n    fig.suptitle(\n        \"Same fault_code (LIDAR_SIM), different root cause\",\n        fontsize=13, fontweight=\"bold\",\n    )\n    plt.tight_layout()\n    plt.savefig(\"../docs/fleet_comparison.png\", dpi=120, bbox_inches=\"tight\")\n    plt.show()\n    print(f\"Left: {seq_a}\\nRight: {seq_b}\")\nelse:\n    print(f\"Only {len(lidar_sequences)} LIDAR_SIM sequence(s). Run the fleet variant for comparison.\")",
+   "metadata": {},
+   "execution_count": null,
+   "outputs": []
   }
  ],
  "metadata": {