Dear friends,
MI4People had a very good start in 2022. Despite the usual challenges of recently born startup NPO, we could achieve progress in many realms:
The number of volunteers that support us is increasing strengthening our operational potential and stability,
We continuously finding new partners in the non-profit sector who are interested in application of Machine Intelligence (MI) to fulfil their missions and we expect many interesting and important projects to be kicked off soon,
And our ongoing social media campaign “Faces behind MI4People” has turned out to be extremely successful: it has already organically reached (i.e., without any paid advertisement) several thousand people and helped us increase public awareness for the idea of MI for Public Good!
But still, there is an enormous amount of work in front of us, both on project and administrative side. And this work cannot be done without your financial support.
On this front, we also have one encouraging news: The Sparkasse Landshut, our bank, is supporting us with their doubling campaign. I.e., it will double your donation up to 100€ if you donate from Monday, February 14, 2022, 9:00 a.m. to Friday, February 18, 2022, 11:59 p.m. (CET) via our donation campaign on betterplace: https://www.betterplace.org/en/projects/103032-mi4people-empowering-public-good-with-machine-intelligence.
Please take advantage of this opportunity to show your support for MI4People.
Your donations will help us create novel MI-applications for Public Good so that great humanitarian causes can also get support from high-tech. We will use your financial contribution to hire dedicated data scientists and engineers and to cover some of our operational costs (such as cloud computing).
We remain very grateful to you for your support and wish you a happy Valentine’s Day!
Let us together make the world a better place for all of us.
Your MI4People Team
News
In line with today’s theme, in this issue of our newsletter we will talk about love, more precisely about love for your mother earth, the environment we all live in, and the creatures who share this beautiful planet with us.
Protecting Environments at Risk with AI
Recently, Capgemini has announced its pro bono project together with the Nevada chapter of The Nature Conservancy that is focused on conservation of the beautiful Mojave Desert in the Southwestern United States.
This desert is a very sensitive ecosystem that is very difficult to monitor because of its size. At the same time, for many people it is an increasingly attractive destination for recreation, such as off-road dirt biking. While this hobby can be helpful for Mojave Desert to increase awareness of people about this unique landscape, the usage of unauthorized routes might lead to substantial damages of this ecosystem. Because of the enormous size of Mojave Desert (nearly 47 thousand square miles), it is very difficult to timely identify the use of unauthorized off-highway vehicles across the landscape. And this is where Artificial Intelligence comes into play.
Capgemini helped The Nature Conservancy to create a tool that uses AI and Machine Learning to analyze satellite imagery and identify trails created by off-road vehicles. These insights are then connected with existing datasets that show the living spaces of at-risk species or areas of human spiritual significance to enable rangers to nip harmful leisure activities in the bud.
Using Machine Learning to fight poaching
The Zoological Society of London (ZSL) partners with Google Cloud on developing a machine-learning-based system that should protect endangered animals from poachers.
Poaching is one of the major risks to the survival of endangered species. The trading of wildlife has a value of between 7 and 23 billion dollars each year, making it the fourth most lucrative crime in the world, after the trafficking of drugs, humans and arms.
Already in 2018, ZSL began to research whether audio monitoring systems might help tackle the illegal hunting problem. They deployed 69 acoustic recording sensors in a nature reserve in Cameroon in order to collect audio data and to analyze how helpful this would be in identifying poaching activity in this area. Thus, ZSL has generated a large amount of data (ca. 350 GB) that is too large to be analyzed by humans.
It is where Machine Learning (ML) can help. ZSL researchers have used a pre-trained, open-source, ML model called YAMNet that was originally created at Google. This model could create an initial classification of all 350GB worth of data within only 15 minutes and identify 1,746 audio instances of being gunshots with high confidence.
From such analyses, the identified suspected audio clips can be analyzed by researchers manually to provide insight on the applicability of audio monitoring against poaching.
AI to minimize waste
Our current consumer society produces enormous amount of waste. A big part of it is the various packaging used for shipping by online retailers. In 2020, ca. 130 billion consumer parcels were shipped around the world and this figure is expected to double within the next five years. A considerable amount of these parcels were sent by Amazon, a company that is responsible – according to some estimates – for producing around 465 million pounds of plastic packaging waste yearly.
But Amazon also tries to shrink its environmental footprint and uses AI to achieve this. It created and deployed an AI-system that uses product descriptions, product images, and certain structured data as inputs to automatically generate decisions on how a particular product should be packed for shipping. The development of this system took more than six years and helped the online retailer cut packaging waste equivalent to over 2 billion shipping boxes.
While saving 2 billion shipping boxes seems to be a drop compared to 130 billion parcels, it is definitely a step in the right direction and shows how MI-technologies can contribute to environmental protection.
Concluding Remarks
MI-Technologies can be broadly used to protect our planet and its biodiversity. You can learn even more examples of how MI already contributes to it in our dedicated articles on environmental protection and wildlife conservation. But it is only the very beginning of a longer journey, and many new MI-applications must and will be developed to save the Earth.