A head-mounted display has a power problem that is structurally different from the same problem on a vehicle or a fixed installation. The load is moving. The operator cannot plug in. The mass budget for batteries is tight, because every gram on the head degrades operator comfort and accelerates fatigue. And the system has to survive a full duty shift, which in practice means ten to fourteen hours from power-on to power-off, not the eight that an optimistic product sheet might assume.
The HMD-1 power architecture is being designed around that constraint from the start, not retrofitted to it.
The load envelope. The active draw on HMD-1 under operational conditions is being measured across three states: display-on at full brightness, display-on at reduced brightness, and display-off with sensor suite active. The third state matters because Sentinel is designed to keep the sensor suite running continuously during a deployment even when the display is dimmed or blanked. The operator needs to be able to bring the display up without a warm-up delay. Power-cycling the display to save power is not acceptable if it introduces a three-second lag at an inconvenient moment.
Preliminary load estimates from component datasheets put the full-brightness draw at approximately 800mW continuous. Reduced brightness is estimated at around 500mW. Sensor-suite-only state is estimated at around 150mW. These are datasheet figures, not measured figures, and the lab treats them as upper bounds until REV A02 board-level power measurement is completed.
Battery sizing. Working from 800mW worst-case continuous and a fourteen-hour window, the energy requirement before efficiency derating is approximately 11.2Wh. Cell voltage on the selected chemistry is 3.7V nominal, giving a capacity target of around 3000mAh before derating. Actual capacity will need to be higher to account for chemistry efficiency, connector and regulator losses, and end-of-life capacity degradation.
The lab is currently sizing the primary cell at 4000mAh with a secondary reserve cell at 1200mAh. The secondary is not a backup in the conventional sense. It is intended to carry the sensor suite through a primary cell swap in the field without dropping the display or requiring a full reboot sequence.
Cell swap. Field-swappable cells are a hard requirement. A system that requires tool access or return to base for a battery change fails the operational envelope. The HMD-1 cell bay is being designed for tool-free access with one hand, accommodating gloved operation. The target swap time is under thirty seconds from unclip to operational.
This creates a mechanical design constraint: the cell bay has to be secure enough not to release under physical stress, accessible enough to operate one-handed with gloves, and sealed well enough to meet the ingress rating. All three at once. The REV A03 housing design is addressing this as a primary constraint, not a secondary one.
What is not being designed in. USB-C charging during operation is not in the current scope. The risk is operator behaviour: a system that can charge from a battery bank while in use will be used that way, often with cable management that creates snag hazards. The lab prefers to keep the charging workflow strictly separate from the operational workflow. Charging happens at base. Deployment happens on cell power.
Wireless charging is similarly out of scope. The addition of a receive coil and charge controller adds mass and a failure mode without adding capability the lab considers necessary.
The power architecture will be published in full as part of the REV A02 documentation package. This note is a summary of where the design currently sits and the constraints driving it.
RELATED PROGRAMME
Sentinel HMD-1 →