Anything that can go wrong will go wrong

Yesterday

We had a very eventful day today. The day began well enough with us getting the bot in water before noon to perform our first profiling mission. Rachel was “tending” the fiber optic cable. This cable forms the communication link between mission control and the robot’s on-board computer. Tending the cable involves being aware of the robot’s current location with respect to the melt-hole and having an approximate estimate of the length of cable that must be spooled out, whether it should currently be spooled in (when the robot is moving towards the melt-hole) or spooled out (when the robot is moving away from the melt-hole) and the direction it should be pulled in/let out.

After the first few grid points, Rachel and Bart noticed that the fiber optic cable was caught somewhere because it wasn’t getting spooled in even when the bot was driving towards the melt-hole. We made the robot retrace its path in the hope of untangling the fiber. The fiber got unsnagged and we move on to further grid points. We had entanglement twice more and ended up driving about 500 m to just entangle the fiber.

Finally, after we had unsnagged the fiber, we decided to profile some more grid points. Just when the robot was about to approach a grid point, we lost communications. In a short time, we regained comunications. The main on-board computer had rebooted. However, the navigation instruments that provides estimates  of pose and orientation were not communicating.  We rebooted the on-board computer in the hope that ths will fix things, but after reboot none of the processes would run. We had no way of bringing the bot back without control of thrusters an knowing the position estimated.

The fiber optic cable can withstand a tensile force upto 300 lbs. The robot is positively bouyant and floats up to sit against the ice when its not being controlled. Ultimately, we ended up pulling the bot back with the fiber optic cable.  The day ended with several possible explanations, but no firm conclusion as to the cause of the various problems. We had a long day and decided to debug the problem the next day.

IMG_7738

Rachel tending the fiber-optic cable.

IMG_7741

John Priscu arrived a few days back. John and Chris discussing the profiling data.

IMG_7743

Bill and Bart pulled the bot about 150 m with the fiber-optic cable.

IMG_7744

The robot back in the melt-hole.

Advertisements

2 Responses to Anything that can go wrong will go wrong

  1. Aniket says:

    Sounds like quite a day!

    Since data was being sent back over the fiber optic you must have had a rough idea of where the robot was (the last pose estimate), right?

  2. sgulati says:

    Indeed, it was. We did know where the robot was – luckily we had lost comms just when we had arrived and were about to stop at a known waypoint. In addition, Bill and Vickie track the robot using a radio beacon, so we know (to within 20 cm) where the robot is under-ice. Still, if we lose control, recovery would mean melting another hole and could take up to 10 days.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: