stats: Axelera Metis board temperature poller

The card exposes no utilization counter usable next to a live detector, so
the poller reports the board-controller temperature read via axcmd (cached
5s; the read must not race detector init).

frigate/stats/hardware.py

@@ -21,6 +21,7 @@ from frigate.detectors.hardware import DEV_ROOT, hardware_prober
 from frigate.util.services import (
     get_amd_gpu_stats,
     get_axcl_npu_stats,
+    get_axelera_board_temp,
     get_cpu_stats,
     get_hailo_temps,
     get_intel_gpu_stats,
@@ -195,6 +196,18 @@ def _poll_axengine(config: FrigateConfig) -> HardwarePollResult:
     return HardwarePollResult(npu={"axengine": npu_usage})
 
 
+def _poll_axelera(config: FrigateConfig) -> HardwarePollResult:
+    # the Metis card exposes no utilization counter without a broker that
+    # cannot run alongside a live detector; the board controller temperature
+    # is the one honest readout
+    board_temp = get_axelera_board_temp()
+
+    if board_temp is None:
+        return HardwarePollResult(ok=False)
+
+    return HardwarePollResult(npu={"axelera": {"mem": "-%", "temp": board_temp}})
+
+
 POLLERS: dict[str, Callable[[FrigateConfig], HardwarePollResult]] = {
     "nvidia": _poll_nvidia,
     "jetson": _poll_jetson,
@@ -204,6 +217,7 @@ POLLERS: dict[str, Callable[[FrigateConfig], HardwarePollResult]] = {
     "rpi": _poll_rpi,
     "intel_npu": _poll_intel_npu,
     "axengine": _poll_axengine,
+    "axelera": _poll_axelera,
 }
 
 
@@ -379,6 +393,8 @@ class HardwareStats:
                     names.add("rockchip")
                 elif spec.detector == "axengine":
                     names.add("axengine")
+                elif spec.detector == "axelera":
+                    names.add("axelera")
                 elif spec.detector == "tensorrt":
                     names.add("jetson")
                 elif spec.detector == "openvino":
@@ -468,5 +484,11 @@ def get_hardware_temperatures(detector_type: str) -> list[float | None]:
     elif detector_type == "hailo8l":
         hailo_temps = get_hailo_temps()
         return [hailo_temps[name] for name in sorted(hailo_temps.keys())]
+    elif detector_type == "axelera":
+        # Metis reports one board temperature through the SDK's axcmd tool
+        board_temp = get_axelera_board_temp()
+
+        if board_temp is not None:
+            return [board_temp]
 
     return []

frigate/util/services.py

@@ -831,6 +831,70 @@ def get_axcl_npu_stats() -> dict[str, str | float] | None:
     return None
 
 
+_axelera_board_temp_cache: tuple[float, float] | None = None
+_AXELERA_BOARD_TEMP_TTL = 5.0
+
+
+def get_axelera_board_temp() -> float | None:
+    """Get the Axelera Metis board temperature through the SDK's axcmd tool.
+
+    The read goes through the board controller and does not open an inference
+    stream, so it is safe to call while a detector is running. ``axcmd`` ships
+    with the runtime wheel; ``find_tool`` prefers the runtime-installed copy
+    under the user base. Results are cached for a few seconds so one stats tick
+    spawns axcmd at most once.
+
+    NOTE: this function is blocking (up to ~5s on a hung board).
+    """
+    global _axelera_board_temp_cache
+
+    now = time.monotonic()
+    if (
+        _axelera_board_temp_cache is not None
+        and now - _axelera_board_temp_cache[0] < _AXELERA_BOARD_TEMP_TTL
+    ):
+        return _axelera_board_temp_cache[1]
+
+    try:
+        from frigate.util.runtime_deps import find_tool
+
+        axcmd_path = find_tool("axcmd")
+    except Exception:  # noqa: BLE001
+        axcmd_path = shutil.which("axcmd")
+
+    if axcmd_path is None or not os.path.exists(axcmd_path):
+        return None
+
+    try:
+        result = sp.run(
+            [axcmd_path, "--board-temp"],
+            capture_output=True,
+            text=True,
+            timeout=5,
+            check=False,
+        )
+    except (OSError, sp.TimeoutExpired):
+        return None
+
+    if result.returncode != 0:
+        return None
+
+    # anchor on the board-controller line ("N: temp_sensorX@YY = <value>") so
+    # unrelated "= <number>" output (firmware/version preambles) is not picked up
+    match = re.search(r"temp_sensor[^=\n]*=\s*([\d.]+)", result.stdout)
+
+    if match is None:
+        return None
+
+    try:
+        temp = round(float(match.group(1)), 1)
+    except ValueError:
+        return None
+
+    _axelera_board_temp_cache = (now, temp)
+    return temp
+
+
 def try_get_info(f, h, default="N/A", sensor=None):
     try:
         if h: