I've created a folder ROV_power_consumption that includes files, code, and figs centered around understanding ROV power consumption. The motivation is to develop the ROV power cosumption analyses / visuals that will be used in our first field methods manuscript - we are especially interested in comparing/contracting the various ROV development versions, e.g., battery-powered vs surface-supplied power.
The scripts wrangle.R and visualize.R both use source() to pull in functions from functions.R Input data files can be found within data, and the resulting .csv output files are in results.
Example output from wrangle.R can be found at V3_power_consumption.csv, with columns for volts, amps, watts, log10 watts, Ah, mAh, and Wh. The column transect contains a label for any real-world transect, e.g., "1" for transect 1, and defaults to "0" if the row associates with the ROV flying to or from a transect.
With all this in place, we can look at a variety of aspects of power consumption.
For example, from an older telemetry file back when Nereo was battery powered (our "V1" configuration in the manuscript), i.e., it had a DVL, integrated Blue Robotics lights (no external battery powered lights), and had no payload skid, we can see the following from surveys in Elliott Bay in calm conditions (thrusters at 20-30%):
Ah consumed across time:
Watts consumed across time:
and finally, Watts log10(x+1) transformed, but with the axis text exponentiated such that the labels are on the natural scale, and the transects are depicted as stacked kernel densities:
One big takeaway is that the V1 ROV typically consumed less than 100W when surveying - that's good, as that leaves plenty of power available for, e.g., the integrated DeepSea lights that we are working on now at Seattle-Aquarium/CCR_development#34.
Several items remain:
I've created a folder ROV_power_consumption that includes files, code, and figs centered around understanding ROV power consumption. The motivation is to develop the ROV power cosumption analyses / visuals that will be used in our first field methods manuscript - we are especially interested in comparing/contracting the various ROV development versions, e.g., battery-powered vs surface-supplied power.
The scripts
wrangle.Randvisualize.Rboth usesource()to pull in functions fromfunctions.RInput data files can be found withindata, and the resulting .csv output files are inresults.Example output from
wrangle.Rcan be found at V3_power_consumption.csv, with columns for volts, amps, watts, log10 watts, Ah, mAh, and Wh. The columntransectcontains a label for any real-world transect, e.g., "1" for transect 1, and defaults to "0" if the row associates with the ROV flying to or from a transect.With all this in place, we can look at a variety of aspects of power consumption.
For example, from an older telemetry file back when Nereo was battery powered (our "V1" configuration in the manuscript), i.e., it had a DVL, integrated Blue Robotics lights (no external battery powered lights), and had no payload skid, we can see the following from surveys in Elliott Bay in calm conditions (thrusters at 20-30%):
Ah consumed across time:
Watts consumed across time:
and finally, Watts
log10(x+1)transformed, but with the axis text exponentiated such that the labels are on the natural scale, and the transects are depicted as stacked kernel densities:One big takeaway is that the V1 ROV typically consumed less than 100W when surveying - that's good, as that leaves plenty of power available for, e.g., the integrated DeepSea lights that we are working on now at Seattle-Aquarium/CCR_development#34.
Several items remain: