AI-Enabled Antarctic Sea-Ice, Iceberg Trajectory, and Navigation Decision Support System

The receipts 🧾

Every red flag on this statement, in full. These are the four places it bites.

1. Exhibit A

   Three distinct modelling problems in a single sentence is a scope trap — a team that splits evenly across forecasting, drift and routing will do all three shallowly, and the forecast is the one the other two depend on

2. It gets worse

   With 277 characters of specification you are defining the deliverable yourself, and a panel that reads the sentence differently will measure you against a system you did not build

3. Still reading?

   Ice-aware routing needs a cost model for how hard a vessel finds a given ice concentration, and without a vessel ice class and resistance model your route optimises a difficulty you invented

4. And the finisher

   Sea ice forecasting skill degrades sharply beyond a couple of weeks, so pick a horizon you can actually demonstrate skill at and report against persistence, which is a surprisingly strong baseline

The damage report 📉

Every score this statement earned, and what each one actually costs you.

• Feasibility

  4/5

  Actually buildable, which on this slate is rarer than it sounds. Do not squander it on scope.

  The data situation here is unusually good — daily passive-microwave sea ice concentration reaches back to 1979, Antarctic iceberg positions are tracked and published, and reanalysis wind, current and temperature fields are free, so both the forecast and the drift model can be trained and back-tested against genuine long records rather than simulated ones.

• Innovation scope

  5/5

  There is something genuinely new here. Do not bury it under another dashboard.

  The statement is a single sentence naming three tasks with no method, no horizon, no resolution, no vessel and no architecture specified, so every meaningful decision about what the system actually is belongs to you.

• Clarity

  2/5

  Nobody is sure what is being asked, quite possibly including the people who asked it.

  At 277 characters it names three substantial modelling problems and stops — there is no forecast horizon, no spatial resolution, no definition of a safe route, no vessel class and no success criterion, so you are writing the specification as well as building to it.

• Acceptance potential

  4/5

  Strong footing before you have written a line. Try not to waste it.

  Genuinely underrated — the satellite ice record and the iceberg tracking database are exceptional free resources with decades of history, sea-ice forecasting has published benchmarks you can position against, the visual is memorable, and polar navigation is a domain essentially no team will enter.

• Effort

  Heavy

  Heavy. Somebody on this team is not sleeping in week three. Pick who, on purpose.

  A spatiotemporal ice forecast, a physically grounded drift model and an ice-aware routing engine are three separate problems in three different disciplines, each with its own data pipeline and validation story, plus a map interface tying them together.

• Demo-ability

  Easy

  Easy to demo — and so is everyone else's. Working is the floor here, not the achievement.

  A polar map with a forecast ice field, drifting bergs and a ship's track threading between them is striking and self-explanatory, and holding out a real date so the forecast can be checked against what actually happened turns the visual into evidence.

• Data

  None supplied

  No dataset comes with this one, so every accuracy figure you quote is a number about labels you invented.

  Nothing is provided with the statement. You are sourcing, cleaning and labelling it yourself, and that work is invisible in the demo but very visible in the questions.

The demo they will have already seen 🪞

Somewhere around 150–360 teams are heading here, and the description is doing the choosing for most of them. They will read the same brief, reach the same architecture, and build a version of the same demo you are planning. Being correct is the floor. If your five minutes could be swapped with the team before you and nobody in the room would notice, you have not picked badly — you have built predictably, which costs exactly the same and hurts more.

What survives 🛡️

The ground worth standing on when the questions start.

• The satellite sea ice record runs continuously since 1979 at daily resolution, which is one of the best free training datasets available for any statement on this portal
• Antarctic iceberg positions are tracked and published historically, so your drift model can be validated against real trajectories rather than assumed physics
• Published sea-ice forecasting benchmarks exist, so you can position your skill against a documented baseline instead of defining your own standard