AI For Good: Startups are using machine learning and computer vision for health and disaster relief
Looking at startups that are leveraging data science, artificial intelligence, and drone technology to solve tough challenges in aid, health, and natural disaster response as they occur in developing countries
Artificial intelligence has transformed how we make use of vast amounts of data. Yet, most headlines seem to focus on how AI is reshaping industries and less on its humanitarian applications. In many developing countries, startups have started to use drones, machine learning, and computer vision to tackle some of the most pressing problems faced by local governments and NGOs. In certain cases, this type of top-level tech has proven to be useful to compensate for a lack of resources in health, infrastructure, or to prioritize humanitarian aid.
Let’s dive into 3 examples of startups that are using AI and machine learning to improve the living conditions in developing countries.
Zipline: “Lifesaving Drone Delivery”
Launched in 2011, Zipline is a drone delivery startup that was created in San Francisco. However, its core business is to deliver lifesaving medical supplies and blood units in Rwanda. As it relies on the latest UAV or unmanned aerial vehicle technology, the teams at Zipline are able to overcome the country’s poor road infrastructure and reach populations in remote regions. To date, the startup has delivered more 7,000 units of blood over 5,000 autonomous flights, successfully saving hundreds of human lives.
So, what’s next for Zipline? The company recently released their latest drone model, deemed the “fastest delivery drone in the world”, and carrying up to 1.75 kg of cargo. Zipline’s CEO Keller Rinaudo stated they are working with more countries such as the United States and Tanzania to launch their drone delivery service in more regions: “We’ve been hard at work to improve our technology and are ready to help save lives in America and around the world.”
WeRobotics: “Robotics for the Benefit of All”
Created in 2016, WeRobotics aims to scale and localize the positive impact of robotics and AI across social good sectors such as aid, health, and agriculture. They have already launched robotics projects in over 9 countries, carrying out drone missions to fight crisis such as the Zika virus in Peru. To employ theses tools at scale, WeRobotics created “Flying Labs”, which are localized training programs to enable local partners in low-income countries on how to use their drones in an efficient way.
Earlier this year, they initiated an Open AI Challenge in partnership with the World Bank and OpenAerialMap to address destruction by cyclones, tsunamis, and other natural disasters in the South Pacific. The challenge was to build an effective machine learning classifier to quantify trees, which are an important source of food security in such situations. Participating teams were provided with satellite data and UAV imagery, then asked to build predictive models on top of it to automate feature detection. The winning team from Simon Fraser University (Canada) presented a classification model that automates the identification of trees, and an interface that overlays detections with confidence levels.
One Concern: “Benevolent Intelligence”
Another organization, One Concern, has created a revolutionary AI system to better predict the impact of natural disasters. Founded by Ahmad Wani, the original idea behind it was to bring science behind disaster response by making sure first responders have the necessary information to know where to focus their efforts. The One Concern platform uses geological data from various sources and adds layers of machine learning to predict damage with great details, down to individual city blocks.
Since its funding in 2015, One Concern has raised more than $22 million in Series A. Now, the company is on its way to build more AI solutions for wildfires, floods, and hurricanes response. According to his post, Wani has fully embraced the vision of building “Benevolent AI” to fight the impacts of natural disasters, climate change and overcome growing inequality around the world. Something that can only be praised as the number of disasters worldwide has quadrupled since 1970.
Conclusion: Startups are ideally positioned to build AI for Good
Technologies as impactful as AI and machine learning need to be developed within certain boundaries, one of them should be to provide assistance and support to people in need. What startups like Zipline, We Robotics, and One Concern are doing with AI technology defines the humanitarian side of the field. Aside from having a world impact for populations in developing countries, it helps building a more sophisticated society.
The Quest For Ethical AI (QEAI) is a weekly publication that aims at uncovering the champions, news, and methods in the space of ethical or responsible AI.
Leo Bancou is a content strategist with a passion for artificial intelligence and its impact on society. Connect with me on LinkedIn
