Power Saving¶

This page collects the practical levers a Slocum pilot can pull to minimize energy use — both to stretch a normal mission and to keep a glider alive as long as safely possible while it waits for recovery (after a weight drop, a leak abort, a schedule slip, or just a long pickup window). It covers saving power while diving and while drifting, and on the surface as well as below it.

Where the energy goes¶

Roughly, a Slocum spends energy on four things:

1. Buoyancy pump — by far the biggest mover; every inflection costs pump work, and more drive (cc) and faster dives cost more.
2. Pitch (battpos) motor — repositioning the battery to set pitch.
3. Science — the science computer and sensors (CTD, optode, fluorometer, altimeter, etc.).
4. Comms and "housekeeping" — Iridium/Freewave, GPS, the flight computer's wake cycle, and time spent awake at the surface.

Saving power while diving (mission design)¶

These are the levers you set in the .mi/.ma files for a normal endurance mission. Stacking several of them is what gets a glider to the low A·h/day numbers above.

Trim science¶

• Turn science off entirely (put c_science_on 0, put c_science_all_on_enabled 0) for the biggest base-load cut, or
• Subsample: sample only every X downcasts, or sample on the downcast only. Reducing what's in sbdlist.dat also drops the data volume you transfer at the surface.

Use Low Power Mode¶

Low power mode reduces the wake/cycle time of the flight (and science) computers so the glider effectively sleeps between the things that actually need attention. Enable it from the start of a mission for the most benefit, though it still helps near the end of a battery.

sensor: u_low_power_cycle_time(sec) 30   # > 0 enables LP; 30 is the recommended max
sensor: u_science_low_power(sec)   30    # science computer powers down between samples
sensor: u_alt_reduced_usage_mode(bool) 1 # let the altimeter sleep when not needed
sensor: u_alt_min_depth(m) <as deep as comfortable>

• u_low_power_cycle_time only needs to be set once; it is not reset during the mission. Recommended max value is 30.
• The altimeter blocks low power. While the altimeter must be on, the glider will not enter low power. In deep, stable water, either turn the altimeter off or push u_alt_min_depth as deep as you safely can so the altimeter is needed for less of each dive. (Set it very deep during simulations.)
• Monitor whether it's actually working with x_low_power_status (it tells you why low power is disabled — altimeter on, GPS on, motors moving, surfacing, inflecting, etc.) and x_cycle_time. For science low power, watch m_science_on.

Fly the buoyancy and pitch motors gently¶

• Use autoballast. TWR highly recommends autoballast on G3 specifically for energy conservation (and to avoid unnecessary speed). With 8.0+, default flight drive is ±260 cc; using the full HD drive (toward ±1000 cc) "significantly increases energy consumed" — and legacy ±1000 cc missions can cause fast, steep, uncontrolled dives in shallow water.
• Dive deep and slow. Fewer inflections per unit of data; minimize the buoyancy drive used per yo.
• Shallow pitch angle. Operators fly endurance missions as low as ~18° (vs. a typical ~26°) because the smaller angle reduces pitch-motor work — as long as you're not getting pushed around more than you like.
• Fixed battpos / fixed bpump. If you know settings that work, turn servo mode off on the pitch motor and use a fixed battery position (and a fixed buoyancy value) so the motors aren't constantly hunting.

Minimize surface time and comms¶

• Lengthen the surfacing interval — surface for fewer reasons. You can comment out "surface for waypoint" and surface only periodically for a GPS fix and data transfer (see Drifting below the surface).
• Spend as little time on the surface as possible transferring files (smaller sbdlist, fewer files).

Drifting below the surface (waiting submerged)¶

If you can afford to be underwater, drifting at depth ("bathtub" mission) is usually the lowest-power way to hold station and avoid being swept around on the surface.

• Set a target drift depth and a wide deadband so the buoyancy engine stays off most of the time. If the depth/angle deadband is tight and the glider has to correct constantly, drift-at-depth can actually use a lot of energy; if the deadbands are open, usage is reasonable. Example field starting point: target_depth = 100, target_deadband = 150 (the band spans the surface so the glider rarely needs to pump).
• Hours of drift per segment is fine; the practical limit is more environmental/operational than a hard number. The glider can also drift nose-up (e.g. ~20° for upward-facing radiometers).
• Combine with low power mode and science off/subsampled for the lowest draw.
• For holding a position (virtual mooring), pick a number of dives between GPS surfacings, set num_half_cycles_to_do accordingly, and surface for time every few hours just to transfer and re-fix.

Drifting on the surface (waiting for recovery in GliderDOS)¶

When the glider can't dive (weight blown, leak) or you want it parked at the surface, the goal is to stay in GliderDOS, call in on a schedule, and shut off everything you don't need. This is the "standard low-power drift" recipe.

Stretch the timers and call in less often (values are examples — set to suit your battery and pickup window; 1–6 h callbacks are typical):

put u_iridium_max_time_til_callback 3600   # up to 1800 s is the documented max for c_iridium_time_til_callback
put u_max_time_in_gliderdos 3600           # how often it cycles into a mission to try to call in
put c_science_on 0
put c_science_all_on_enabled 0
put c_console_on 0                          # turns off Freewave console

Then take non-essential devices out of service (use - <device>), as appropriate for your vehicle — e.g.:

GPS · pinger · attitude / attitude_tcm3 / attitude_rev · ocean_pressure · pitch_motor · science_super · fin_motor · digifin · altimeter · thruster

…and run a callback script at your chosen interval.

Staying out of a mission: lastgasp and reverting settings¶

Keeping a stuck glider in GliderDOS is fragile — the firmware fights you:

• Settings silently revert. Operators report that within a day, u_max_time_in_gliderdos reverts to 900, and c_science_on / c_science_all_on_enabled flip back to 1, after which a long callback lets the glider sequence into initial.mi / lastgasp.mi and possibly dive or abort.
• A reported workaround to keep it from sequencing is put u_max_time_in_gliderdos -1 (stay in GliderDOS indefinitely). Re-check your settings every call.
• lastgasp.mi is the minimal survival mission the glider falls into; like initial.mi it does not use the 8.0 default buoyancy drive changes. Note that you cannot take a critical device (e.g. Freewave/console) out of service from GliderDOS with the glider on the pier — it's a critical device, not even required in autoexec.mi.
• A simple robust approach used in the field: just increase the GliderDOS and callback timeouts and run an Iridium callback script on a fixed cycle (e.g. ~11 min when close to pickup, longer when not) so the glider keeps reporting position without diving.

Know when to stop: battery shelves¶

Power saving buys time, but the battery voltage shelf sets the real deadline.

• With the older 3S lithium design, TWR recommends you stop diving and drift once energy reaches the first shelf.
• With the 4S design you can keep operating on the first shelf, but the appearance of the second shelf drop may be the last time the glider communicates — do not deplete either chemistry to the second shelf operationally.
• Derate stored capacity (~3%/year is TWR's rule of thumb; field experience suggests partly-used packs can lose more). See the primary battery page for capacity, derating, and the voltage-shelf details.

Quick reference¶