Make tourism more respectful of marine life
Presentation
The subject
Seas and oceans cover
70% of our world and hold 97 percent of all water on the
earth. They link people, supply food and oxygen, manage our climate by
absorbing about a quarter of CO2 created, and support a diverse
range of animals [1]. It therefore seems essential to conserve and
use our oceans in a sustainable way.
One of the major factors impacting marine life is tourism. Today,
tourism is responsible for
8% of global greenhouse gas emissions, contributing to
ocean warming and acidification [2]. In addition to these effects,
there are other direct pressures such as pollution, destruction of
habitats and increased coastal development.
The tourism sector must therefore engage in working to protect the
seas and oceans for the benefit of future generations. Today,
long-term tourism potential depends on
developing sustainable practices and ensuring that human
activities have minimal impact on ecosystems [3].
Technological module : Artificial intelligence
« In general terms, AI refers to a broad field of science
encompassing not only computer science but also psychology,
philosophy, linguistics and other areas. AI is concerned with
getting computers to do tasks that would normally require
human intelligence. » Deloitte
In everyday life, Artificial Intelligence is particularly present
and is applied in various fields thanks to machine learning.
Machine learning will involve working with small to large datasets
by examining and comparing the data to find common patterns and
explore nuances. Today there is a real development in this field,
so we might as well exploit it to solve the problem of tourism on
the marine life.
Project 1 : Happywhale
Project carriers : The two founders, Ted Cheeseman and Ken
Southerland
Beneficiaries : The research community but also anyone
interested in marine mammals
Users : Whale enthusiasts and tourists who want to enrich
the database of whales by taking and uploading pictures during
whale spotting tours
Need : Increment worldwide understanding and caring for
marine environments by creating high quality content for science
and education
Principle : During whales spotting tours, ecotourists can
take pictures of the different whales and upload them on the
Happywhale platform. The algorithm will collect the data to match
with known whale’s photos from the existing database to give more
information about the specific whale taken in photo (ID number,
sex, tracking position …). If the whale is not register in the
scientific collection, you will even have the possibility to adopt
and name. You can even be notified of her migration around the
world. Happywhale is therefore a collaborative, engaging and
rewarding platform [4] [5] [6].
Main technologies involved : It is an artificial
intelligence that is based on automated image recognition
developed with support from Google’s Kaggle Platform. The
algorithm uses shape, design and highlights in a whale’s tail. The
exactness of this calculation is approximately 97% to 99% of
potential matches.
Project 2 : Refill Now map
Project carriers : The Italian company Refill now
Beneficiaries : The tourists but also the marine
fauna/flora that will be protected from plastic pollution
Users : The tourists who want to reduce their usage of
single use bottles during their trips
Need : Reduce the usage of plastic water bottles by
creating an international network of refill water points easily
findable on Refill Now’s online map
Principle : When travelling, tourists can use the online
platform to find water points to fill their reusable water
bottles. On the map, the user can find fountains, natural springs,
drinking water dispensers, sources but also private areas such as
hotels or restaurants providing access to a tap. By finding easy
access to refill their bottle, tourists find it less necessary to
buy single-use bottles during their trip. Refill Now Map help to
diminish marine plastic contamination from land-based activities
by encouraging and empowering bottle reuse [7] [8] [9].
Main technologies involved : It is an interactive online
map equipped with a navigation system.
Project 3 : I clean my sea
Project carriers : The french oceanographer Aymeric Jouon
Beneficiaries : The marine fauna and flora that will be
cleaned and preserved from plastic pollution
Users : Anyone who wants to help preserve the oceans
Need : Facilitate the collection of marine litter through
new technologies and thus preserve the marine environment
Principle : With this application everyone can report
marine litter on the Clean My Sea application which will then be
collected by the boat "The Collector". Indeed, the sailors will
have access to the photo, its position, and the trajectory
forecast. In addition to this application, the startup also uses
coastal cameras and satellites to identify areas of plastic waste
accumulation [10] [11].
Main technologies involved : The application uses
geo-location to allow the user to report where the marine litter
is located. The location is then sent to the servers and can be
used to send extra support to collect the waste.
Project 4 : The World’s Ocean – Google Earth
Project carriers : Google Earth
Beneficiaries : People who want to discover the marine
environment in the four corners of the world without having to
travel
Users : People who want to discover the marine environment
in the four corners of the world without having to travel
Need : Inspire and educate a new generation of global
citizens to discover the world while raising awareness of how to
protect the environment to keep it intact
Principle : Thanks to Google Earth you can travel,
explore, and learn about the world by interacting with a virtual
globe. Within Google Earth, Oceans lets you visualize the ocean
floor in 3D vision and makes you discover the underwater life.
This one proposes a way to raise awareness by showing "dead zones"
(victims of pollution), protected areas, preferred species, etc
[12] [13].
Main technologies involved : Google Earth is a software
program developed to visualise the Earth with an assembly of
aerial or satellite photographs. It digitally stitches of images
using the highest quality pixels from each photograph to create
the clearest view of the seabed.
The selected project : Happywhale
Carriers and actors of the project [14]
Ted Cheeseman – Co Founder and Director
From an early age, Ted grew up surrounded by animals. Both his parents were naturalists and educators, so it was only natural that in 1994 Ted joined the family business as an ecological safari leader. In the meantime, he also obtained a Master’s degree in Conservation Biology from Duke University. Since then, Ted has been heavily involved in all facets of Antarctic tourism management. For five years he was a member of the executive committee of the International Association of Antarctic Tour Operators (IAATO), where he worked to reduce the environmental impact of Antarctic tourism. In 2015, he decided to launch with Ken Southerland Happywhale.
Ken Southerland – Co Founder and Lead Developer
Ken Southerland holds a PhD in aerospace engineering from the University of Michigan. Prior to founding Happywhale with Ted he created, the a cataloguing system for museums (Gallery Systems) and then launched his startup, called SAMsix, developing a GIS suite that allows utilities to map and perform network analysis. Today, he feels extremely fortunate to be able to use his system architecture skills to help raise awareness and protect marine wildlife.
Marilia Olio and Hayley Newell – Data Manager
As a data manager, their role is to manage the large database of the various whale specimens encountered. In this context, they have to collect and organize the data. These different elements must be able to be easily exploited later by different scientists.
"Cascadia research collective" and "Allied Whale" – Partner Organizations
Cascadia Research Collective , situated in Olympia,
Washington, is a non-profit, tax-exempt scientific and educational
organization.
Allied Whale is participating in a wide range of field
research programs, both scientifically and geographically. Their
biologists and students do research on whales in the Gulf of Maine
and collaborate with American and international institutions
throughout the world. In addition to field research, Allied Whale
maintains the world's largest database of photo-identified
humpback and finback whales.
Research question
How can Happywhale uses the photos collected by tourists during whale watching excursions to provide important insights for the scientifc community to preserve and protect whales ?
The reason I selected this project
I chose the project because I find it original and relevant. I find it exciting and admirable that people are mobilising and committing themselves to help the scientific community about whale research. Indeed, this project can really help the community to track the mammal populations, learn about migrations and even estimate their survival rate. Moreover, the fact that Happywhale uses artificial intelligence and that it is a subject that interests me confirmed my choice.
Users
Users can be anyone, they just need to have a camera and be in an area where whales can be seen. So it can be as simple as tourists who decide to go on a whale watching trip. Other users may be fishermen, sailors, locals or anyone who is in an area where whales live. The element that brings them together is that they are passionate about whales and want to make a positive impact by helping the scientific community with this issue. They want their actions to have an impact.
Key Features
- Submit whales’ photos
- Track whales
- Browse info about whales
- Name and adopt a whale
- Interact with the Happywhale’s community
- Make a donation
Persona
UX Story Board
Technical Analysis
General principle
The company uses an Artificial Intelligence-based automatic picture recognition technology to compare whale photos to find out if the animal has already been identified and if so find the corresponding whale.
Technical presentation of the Artificial Intelligence used
The algorithm was developed with the help of
Kaggle Platform [15]. On this platform, companies present
data science problems and offer a prize to the data scientists
with the best performance. More than 2'120 teams entered the
competition in 2019 to develop the algorithm of whale's
identification for Happywhale.
The co-author of the algorithm, Jinmo Park (Seoul - Korea),
noticed that the process of identifying individual whale images is
quite similar to that of human facial recognition. He decided to
simply modified the ArcFace code that is basically used for facial
recognition. This model takes two facial images as input and
calculates the distance between them to determine the probability
that they are the same person [16]. In the case of whales, the
focus here is on the tail, which have unique features (colours,
spots, shape ...). The matches are then ranked to arrive at a
single proposed correct match for the known whales. The algorithm
is therefore
based on supervised learning and uses classification.
In order for the algorithm to work it is essential that data are
made available to it. The datasets used come from
individuals who have submitted their photos on Happywhale but
also from scientific marine databases. In addition to the photo submission, each person is asked to
indicate the date, location and if the photo was taken on board of
a boat, the name of the shipping company (whether it is for
tourist's excursions or affiliated with a research group)[17].
Indeed, it is necessary to indicate the context of the photo.
One of the biases to be aware of is the possibility that some
individuals are duplicated in the database. Indeed, it can
happen that Happywhale has mistakenly given to the same animal two
separate unlinked identifications. These
"false negatives" are usually the consequence of low
quality photos or whales whose recognised traits have altered
dramatically and for whom no match has been identified.
Added value thanks to Artificial Intelligence
The AI used in this project saves a considerable amount of time in identifying each of the species submited. In fact, scientists can use this time saved to improve their research on whales in order to increase global knowledge, and foster ocean conservation. Indeed, identifying each whale and seeing if it is present in a huge database would be a huge waste of time and exhausting.
Sources used
[1] https://www.wwf.fr/champs-daction/ocean[2] https://www.lemonde.fr/blog/oceanclimat/2018/06/08/un-tourisme-responsable-pour-preserver-locean-et-proteger-le-climat/
[3] https://www.un.org/fr/chronicle/article/le-tourisme-doit-demontrer-son-engagement-preserver-la-vie-sous-marine
[4] https://happywhale.com/home
[5] https://www.outdoorproject.com/articles/meet-happywhale-app-identifies-and-tracks-whales>
[6] https://www.101sustainableideas.com/idea/happywhale>
[7] https://www.refill-now.com/en/how-it-works/
[8] https://www.101sustainableideas.com/idea/refill-now-a-network-and-interactive-map-of-water-refill-stations/
[9] https://www.infoelba.com/island-of-elba/maps/fountains-water/
[10] https://icleanmysea.com/fr_fr/
[11] https://www.francebleu.fr/infos/environnement/i-clean-my-sea-une-application-mobile-pour-signaler-les-dechets-plastiques-en-mer-1600943338
[12] https://earth.google.com/web/data=CiQSIhIgNjZlNzQ1NjM3OGMwMTFlOGJmMTZhMTM2MjYyMmZhYjA
[13] https://www.livescience.com/65504-google-earth.html
[14] https://www.happywhale.com/about
[15] https://www.kaggle.com/c/humpback-whale-identification/overview
[16] https://medium.com/axinc-ai/arcface-a-machine-learning-model-for-face-recognition-5f743cdac6fa
[17] https://arabianseawhalenetwork.org/wp-content/uploads/2020/06/sc_68b_ph_01_flukebook-and-happy-whale-platform-comparison-1.pdf
Sources of photos for the different projects
HappyWhaleRefill Now
I clean my sea
The World’s Ocean – Google Earth
Tourists during an exploration - HappyWhale