Inside Cervest’s plan to map global climate risk

It was five in the morning when Iggy Bassi received a phone call telling him that a $ 6 million mill on his sustainable farm in Ghana, West Africa, had been crushed by high winds. Locals called the destruction an act of God. This was not enough for Bassi.
“I don’t believe in the acts of God,” he said. “There’s the science behind it – we just don’t understand it. ”
Soon after, a flash flood destroyed an entire season’s crops.
“It piqued my interest,” said Bassi, who couldn’t understand why his weather warning systems had given no indication of the carnage to come.
“I was really struck by one problem: that we could never get really reliable climate data, event data or weather data. “
He wondered if artificial intelligence (AI) – especially machine learning – could help. So, at the end of 2015, Bassi sold his Gates Foundation-backed sustainable agriculture business and sought a way to easily make sense of the “minefield” of complex climate data.
“I was horrified to find global science stuck in all of these silos. Yet it is the biggest problem humanity will face in the next hundred years.
Bassi faced a complex challenge. From floods to hurricanes, there are many strains of climate science. All use different types of metrics and scientific approaches, making it impossible to quickly compare and quantify data to get a sense of the real threat.
In other words, comparing precipitation and earthquake risk is like comparing apples and oranges.
“The climate is a problem for everyone”
In 2016, Bassi turned to Dr Ben Calderhead, then head of machine learning at Imperial College London for his statistics program. Calderhead told Bassi that “at the end of the day it’s a math problem.”
To their surprise, no mathematical model capable of pulling together disparate and fragmented climate data existed. The duo began to develop their own machine learning platform for Bassi’s new business: Cervest.
“What we have achieved [is that] someone has to merge all these scientists in the world into one platform, which involved a huge amount of data engineering. And that relies on a huge amount of knowledge in the field, especially around the physical sciences, ”explains Bassi.
The London-based startup has assembled a team of top scientists, engineers, developers and mathematicians, along with Calderhead’s chief science adviser, to help build the platform – known as Earth Science. HAVE.
For two years, Cervest refined Earth Science AI algorithms by feeding it agricultural data, such as the impact of climate on crops. But as the platform’s algorithm improved, Bassi and his team realized there was more they could do.
“What I told myself and my new investors is that the climate is a problem for everyone,” he says. “The climate is not just an agricultural problem. And in fact, whatever we build, we build it in a generalizable way that we can apply to multiple sectors.
It was this realization that led Cervest to extend the reach of Earth Science AI from agriculture to climate risk mapping of every asset on the planet.
Climate risk: overall
The platform uses open source satellite imagery and scientific models to search for “climate signals” such as heat risk and flood risk. For assets, Cervest uses open source data to trace buildings. When data is scarce, such as in rural areas, it uses machine learning to fill in the gaps.
The platform’s AI then assesses the impact of these climate signals and replicates them over different time scales – from days to months, to decades – down to the granular level of a single building.
It can provide alerts when certain thresholds are reached. For example, precipitation may indicate that a drought is likely at a plant in the south of France. These risks can then be linked to other factories to warn of how a weather event will affect a company’s supply chain.
Rather than looking at a narrow side of the climate, Cervest looks at the whole. A building in Miami could be at risk of flooding due to rising sea levels. But what about the increased risk of cyclones from warming oceans? Like a chain of dominoes, climate risk is interconnected.
“That’s the way to think about risk – not through a single lens,” Bassi explains. “Our branch of mathematics allows us not to look at individual climate signals on their own. We see them as a collective whole and get a unified view of everything that is going on.
The mathematical model in question is a Bayesian framework (non-Bayesian parametric, to be precise). In short, it allows the model to adapt its probabilities as the data changes – even with nonlinear data where past models don’t always inform future models.
Earth Science AI does not make firm predictions. He cannot say with absolute certainty that your home will be hit by a hurricane in three years. What it does is create a flow of probabilities that adapt as the input data changes.
“Bayesian allows us to take all of these uncertainties from different sciences and tie them together mathematically to say, ‘we have x percent belief that it’s going to happen to you, but that’s the uncertainty that we also have,'” , says Bassi.
“We need to be fully aware of the uncertainty because there is no perfect prediction in nature.”
Cervest’s business model
For companies looking to spend millions or billions of dollars to build a new factory in a world increasingly susceptible to the volatility of climate change, having the most up-to-date climate risk could prove immensely valuable. For a company operating Unilever-wide, in 60 countries around the world, the benefits could be particularly significant.
Investors seem to agree – the company has so far raised $ 36.2 million, including $ 30 million in a Series A cycle in May 2021.
Cervest plans to offer a freemium model for its Earth Science AI platform. This means anyone can access their dashboard for free and see limited climate risk data for locations around the world. But paying customers – consulting firms, insurers, policymakers – will have access to more granular data and more frequent updates, among other features.
Those who need a more advanced service, such as insurance companies, will also be able to connect to a Cervest API to pull climate data into their own systems.
“People might start using climate not to interpret all these complicated science models, just to use it as a service,” says Bassi. “It’s like a credit score – you just need the result to support a decision. ”
COP26 puts pressure on climate action
In December 2019, then-Bank of England Governor Mark Carney noted that “changes in climate policies, new technologies and growing physical risks will cause revaluations in the value of virtually all assets. financial ”.
Right now, a bank loaning millions to a company to build a new factory will consider credit rating, underlying security, and existing assets, but rarely climate risk.
However, some banks have now developed disclosure guidelines that take climate risk into account. More than 1,000 companies have signed up to the voluntary standard, published by the Climate-Related Financial Disclosures Working Group. But in the next few years, Bassi predicts it will become mandatory – an eventuality that could be a boon to Cervest.
The coronavirus pandemic has created “a real rush for people to reassess their risk frameworks,” Bassi says, which could see companies become more open to climate risk assessment.
The United Nations Climate Change Conference, or COP26, which is held from October 31, added to the focus on reducing climate risks.
“We cannot afford these kinds of events in the future,” Bassi says. “It forces governments and businesses to say, hey, how do we rethink our relationship to other risks? ”
Scientists warn countries must reduce their carbon emissions to zero by 2050 to avoid catastrophic and irreversible damage. There has been progress, but it has been slow. But even if dramatic improvements are made, businesses are still at risk.
“Even if we were to stop all greenhouse gases tomorrow morning, we are already locked in 30 to 40 years with the physical damage already accumulated,” says Bassi.
“So just because you switch to clean energy overnight doesn’t mean your assets are secure. It doesn’t give you climate security overnight.
Ultimately, it could be the balance sheets that push companies to take concrete action against climate change.
“The financial markets realize every now and then that something goes off, that something goes wrong in someone’s assets,” Bassi explains.
“There is a forest fire, there is a storm, there is a flood, there is an extreme weather event. What they realize is that the climate is now having an impact on their yields. ”
This is an updated version of an interview that first appeared in the September 2020 issue of the sister title Verdict Magazine.