How would you map the ocean?

Question

How would you map the ocean?

Asked at Google over a year ago

2.9k views

Asked at

2.9k views

How to answer Technical questions

Interview Guide

Top Technical interview questions

Imagine you're the product manager for Facebook Marketplace. Since many sellers don't mark items as sold, what existing functionality and metrics could you use to determine whether an item has likely sold?7 answers | 20.9k views
What happens when you enter a URL in your browser?6 answers | 10.8k views
How does TinyURL work?5 answers | 317k views
See Technical PM Interview Questions

Invite members
Invite by email

Add another

solders15 · Answer 1 · 2018-08-12T22:08:09+0000

Assuming the Interviewer is from Google.
Interviewer (I): How will you map the ocean?
Me (M): What do you mean by mapping? Do you want to map the floor of the oceans?
I: Yes, among other things.
M: Other things could be mapped include : flora, fauna, volcanic vents, sunk objects, minerals etc.?
I: Yes.
M: What are the goals for Google to map the ocean?
I: Just as we have mapped land, we want to map the oceans too.
M: Do you have any time frame for this to be complete?
I: We would like to give ourselves 10 years.
M: I know you have said oceans, but do you want to map ALL water bodies (seas, bays, lakes, ponds etc both natural and man-made).
I: Yes, but we would like to first start with the oceans, seas and bays.
M: Are there any other constraints that I should know about, for example cost?
I: No, at this stage we would like to explore all potential ways. Cost decisions come later.

M: I’m no expert in this area, but these are some of the ways I would attempt to use:
First, I would look at the technologies that have been previously used to see whether I can reuse any of them.

The approaches:

Ships: Sonar equipped ships – Advantage lesser cost. Better granularity. More time consuming. Probably effective only in shallower parts of the ocean and seas. Affected adversely by surface weather.
Submarines: Use sonars and lasers etc to map the base of the ocean. Use other sensors to probe for material composition. High cost of acquisition, potentially governmental regulations involved, high cost of operation, and high granularity of detail.
Drones/Planes: Low flying drones: Less expensive. Potentially can cover large areas quickly. Multiple drones can, in parallel, scan large areas of the ocean. Probably less effective in deep parts of ocean. Lower granularity of detail.
Satellites: Potentially least effective. Very high cost.

Based on the above analysis and the time deadlines. I would use a combination of drones/low flying airplanes fitted with appropriate sensors, in conjunction with submarines for finer detail where needed to map the oceans.

I would test the technologies in lakes of various depths to make sure I can fine tune the technology before taking it to deeper oceans.

I would also hire “oceanologists” who will serve as subject matter experts to guide this project from a technical standpoint.

Harsha Bornfree · Answer 2 · 2022-10-19T10:48:40+0000

— Answering this just out of passion for both oceans and mapping —

Let’s clarify scope:

Large bodies: Oceans and seas
Small bodies: Gulfs, bays and lakes

Understand possible objectives.

Navigation
- Shipping
- Military
- Fisheries
- Others (recreational)
Resource mapping
- Minerals
- Oil
- Marine life
Scientific exploration
- Physical : temperature, salinity, current, waves etc
- Biological: coral reefs, breeding grounds, migrations
- Geological: underwater vents, volcanoes, rifts and valleys

So primary themes would be:

Navigation
Mapping
Exploration

Much like regular maps actually but for specialized user-personas. Because this isn’t a design question, we’ll get straight to implementation that addresses the above themes.

Navigation

Source maritime navigational material from existing agencies. These should be plenty.
Monitor and collect historic ocean conditions (currents, waves) from satellite and other earth-based monitoring stations.
Put these together using ML-based methods to suggest navigational routes optimized for efficiency and safety taking into account border restrictions etc.

Mapping (the hard part)

This area is high effort and possibly low-impact in the short term.

Use hyperspectral satellite data and remote sensing to identify POI (points of interest).
- Places closer to shore
- Places closer to major human settlements
- Places with high variance in open sea (something might be interesting)
Depending on the location, use ships and ferries equipped with sensors:
- Sonars (sound-based)
- Hyperspectral cameras (light-based)

Exploration

This area is extremely high effort.

We can use AUV (autonomous underwater vehicles) and ROV (Robot operated vehicles) with very similar sensors and extra sensors.
These would help in mapping biological, geological and physical aspects at varying depths for given latitude, longitude patch.

Build overall map:

POIs with high resolution data
- Map surface geology using reconstructions
- Mineral deposits using spectral fingerprints
- Other aspects from AUV data as described in exploration.
Remaining with satellite data
Account for time being one of the dimensions.
- Lat, Long, Depth, Time

As another answer rightly mentioned, we could pilot and perfect these techniques in smaller bodies of water and then move to oceans.

Hung Nguyen · Answer 3 · 2020-04-17T20:47:01+0000

This is my answer. Strongly encourage giving me feedback. Thank you!

Clarify:

Improve the existing map or create the new one? => Improve
Time constraint? => No
Budget constraint? => As low as possible
Other purpose/goals? => No

Solution

SOLUTION

IMPACT

COMPLEXITY

Location tracing sensor

Create a device which is attached to boats, ships and other maritime means to trace the data.

HIGH

The region cover is large because we can collect data around the world.

The time may be fast because we can cooperate with a huge number of maritime means.

The detail of data is good but not too high. We cannot collect in deep parts of ocean because most maritime means float on the ocean

MEDIUM

The cost for R&D is not too high. Domain knowledge is required and we need cost for equipment to mass produce

The cost for operating and maintenance is medium compared to other solutions

Drone

We will use the drone to collect data of the ocean

MEDIUM

The region cover is high because we can cover a lot from above

The time may be slow because of the limited number of drones and R&D.

The detail of data is not too detailed

HIGH

The cost for R&D is extremely high because the cost for developing the drone is not cheap.

The cost for operating and maintenance is high because we have to control all stuff from controlling to recharging.

To save time we can automate the process, but the cost for it is high.

Solar/Wind boat

We will use the self-drive boat operated by the solar or wind energy to collect the data

HIGH

The region cover is high because the boat is self driven and operated by solar energy so they can move around the world.

The time may be slow because of the limited number of self-drive boats and R&D.

The detail of data is good but not too high. We cannot collect in deep parts of ocean because most maritime means float on the ocean

HIGH

The cost for R&D is extremely high because the cost for developing the self-drive boat is not cheap.

Satellite

MEDIUM

The region cover is very large

The time is fast.

The detail of data is not detailed

MEDIUM

Compared to other methods, using Satellite may be cheaper.

Because we can collect the large amount of data in a short time, so the cost for operation is not high

Submarine

HIGH

The region cover is not large

It takes a lot of time for R&D and collecting.

The detail of data is very high

HIGH

The cost for submarine R&D is extremely expensive.

In-depth domain knowledge about marine is required

Marine Community/News

We will keep track of the marine news from different sources like website or creating the marine community to prompt people update the news or data relating to the ocean

MEDIUM

The data relies on human labor so time is the big problem.

The detail and in-depthness of data is high and precise because it’s discovered by humans not machines.

LOW

We don’t have to spend time and money on R&D.

Conclusion

In conclusion, according to the goals, impacts and complexity, the first solutions we should do are to collect the Marine News, create the Marine Community, and R&D location tracing sensor. Then we can test the solar/wind boat and drone to estimate their effectiveness and cost. We also can use satellites once a year to see the big picture and save the money. Finally, if we want to dig into the detail of in-depth marine life, we can implement the submarine to discover the live under the water much more detailed.