Many people believe that significant innovations in technology are often taken for granted each day until they are so ingrained in our existence that we must have them. Some examples today are GPS navigation and smartphones with cameras. The amazing complexity of what goes on inside these devices makes them seem simple. (Greenbiz.com)
Perhaps that is why so many of us are fascinated by how artificial intelligence and sustainability intersect. Applications are being made possible every day by breakthroughs in image recognition, machine learning, analytics, and sensors.
In many cases, the combination of these new technologies may transform systems we know well, as well as approaches commonly used in the sustainability and environmental communities. This could make them much smarter with less human intervention.
Example – Camera Trap
A good example is the camera trap, which is a routine method used to study wildlife habits and biodiversity, and one that has been supported by many large technology companies. (Greenbiz.com). That is where Wildlife Insights comes into the picture – a collaboration between Google Earth and seven other companies, one of which is Conservation International.
Wildlife Insights is the biggest database of public camera trap images on Earth. It has 4.5 million images that have been mapped and analyzed with AI for such elements as the year, species, country, and more. Scientists can use the system to upload trap photos, visualize regions, and collect insights about the health of various species.
Here’s the amazing thing: This databased driven by artificial intelligence can analyze more than 3 million pictures per hour, compared to 300 or 1,000 a person can manage. Depending on the species, the accuracy ranges between 80% and 98%. What’s more, the system will discount shots where there is no animal present.
The overall effect of Wildlife Insights is faster insights and analysis. This is important as habitats and species are disappearing. The sustainability community needs to move faster and take action quicker. Some believe the time is running out to act on climate change, and 75% of business decision-makers think AI will be vital to devise solutions that enhance environmental sustainability.
We Should Be Cautious About Possible Side Effects of AI
While the future is bright with AI and its role in sustainability, it also is wise to be cautious. Many believe that AI will be key in building trust and making sure data is governed in ways that are reliable and secure. Also, before people get too excited about what AI can do, it is important that it does not make problems worse. This means spending more time looking at ways to make the data centers that drive AI applications use less energy and be less impactful as far as materials used.
Ethically, some experts have two major concerns here. First is that enough energy is used to ensure data behind AI predictions that we rely on is not biased or flawed. This means it will be necessary to spend more time ensuring that many human perspectives are considered and that the numbers are correct from the beginning. Second, it is critical to view these systems as part of the solution, and not a full replacement for human workers.
Sriram Raghavan, IBM’s VP of AI research, has said that new research from MIT and IBM shows that AI will help with scheduling tasks, but it will not have as much of an impact on jobs that need skills such as industrial strategy and design expertise. Workers in 2020 will likely see these effects as artificial intelligence becomes more routine in the workplace around the globe.
Spending for AI Will Skyrocket by 2023
According to projections by the tech market research firm IDC, spending on AI systems may reach $97 billion in 2023, which is 2.5 times more than the $37.5 billion spent in 2019. (IDC.com). This seems to be happening now because of faster chips, better cameras, larger cloud data-processing services, and other innovations.
Five Areas Where AI Will Make a Difference for Sustainability
Where will applications with AI enabled make a difference for corporate and environmental sustainability? Below are five areas where AI should have a major impact over the next 10 years:
- Automated energy management: The system for tracking sustainability data today is an old technology that has been with us on PCs for 40 years. Two early examples of how AI can help with energy management have been made by Google, which uses it to enhance efficiency and the renewables it uses in its many data centers. Another is the cold storage firm Lineage Logistics, which uses AI to make schedules for its dozens of warehouses.
- Enhancing soil conditions and crop yields: Sensors and drones that monitor fields are viewed as a vital part of helping agricultural companies make better decisions about plant nutrition, fighting disease, and hydration. A recent example is a crop emergence solution that was tested by the AI startup Taranis (with John Deere backing), along with the drone spraying company Rantizo. The intelligence driving the system came from Taranis; it uses AI to analyze and monitor aerial imagery.
- Modeling climate risks in the future: A good example is what is being done with AT&T’s work with Argonne National Laboratory. It is using its proprietary database of information about the telecommunications network it has with Argonne climate models to predict how the effects of climate change could affect business operations up to 30 years in the future. It is expected that insurance companies and large financial services firms to invest in improving their analytics capabilities with artificial intelligence.
- Protecting biodiversity: The Wildlife Insights initiative is one example of how experts are using imager and data-analysis technology to get a better idea of how the Earth is changing. Google is powering many of these projects, as is Microsoft, which has its AI for Earth project. One effort it funded at the end of 2019 is Wildbook, an effort between the Wild Me Oregon nonprofit and researchers at the University of Chicago, Princeton, and Rensselaer Polytechnic.
- Checking provenance across supply chains: Many next-gen traceability systems are being tested in different industries from coffee to seafood, and it all comes down to the blockchain. This is a fancy name for electronic ledger technologies. The bottom line here is a lot of machine learning is behind many of these sophisticated applications.