Earthmover: Weather & Geospatial Data Solutions

The Evolution of Earthmover: From Climate Tech to Daily Weather Insights

The sheer volume of data generated by observing our planet from space is immense. However, Ryan Abernathey and Joe Hamman, the founders of a burgeoning startup, quickly discovered that extensive data alone wasn't sufficient for sustained growth.

Consequently, their data-focused climate technology company, Earthmover, underwent a strategic shift in direction.

A Focus on Short-Term Climate Impacts

This transition doesn't represent a complete departure from climate technology. Instead, the company is narrowing its temporal focus to examine the impact of climate on everyday experiences – specifically, the weather.

“The key to a compelling application of our platform lies in data that undergoes frequent changes,” explained Abernathey, Earthmover’s co-founder and CEO, in an interview with TechCrunch. “This is where there’s a greater sense of urgency for effective solutions.”

He further elaborated that this dynamic data is primarily found in areas like weather patterns, wildfire occurrences, and newly acquired observations.

The Static Nature of Long-Term Climate Data

Abernathey pointed out that while climate projections are undeniably significant, they tend to be relatively static, with substantial updates occurring only every few years.

Earthmover’s Core Technology: A Robust Data Structure

The foundational element of Earthmover’s offering remains its sophisticated data structure. This structure is engineered to manage extensive and intricate datasets.

“Within the geospatial field, it’s known as a raster. In the realm of AI, it’s referred to as a tensor. And, historically, in Fortran programming, it was simply called an array,” Abernathey clarified.

Building upon this core, the startup has developed a suite of tools designed to assist clients in extracting valuable insights from their data.

Securing Funding and Gaining Traction

This strategic pivot has enabled Earthmover to attract paying customers – currently exceeding ten – and secure a $7.2 million seed funding round, as exclusively reported to TechCrunch.

Lowercarbon Capital spearheaded the investment, with additional participation from Costanoa Ventures and Preston-Werner Ventures. The funds will be allocated to the development of new tools integrated with the company’s data storage platform.

Leveraging Open Source and Cloud Infrastructure

Earthmover’s architecture is built upon open-source software, including Xarray, Pangeo, and Icechunk. It operates across major cloud providers like Google Cloud, AWS, and Microsoft Azure, as well as on-premise server environments.

Both Abernathey and Hamman, Earthmover’s CTO, possess extensive experience within the open-source community, having contributed to the development of Pangeo and Xarray.

Managing Large-Scale Geospatial Data

The choice of these tools was partly driven by their efficiency in handling Earth observation data, which can rapidly accumulate to terabytes or even petabytes in size.

Abernathey indicated that typical Earthmover clients manage between tens and hundreds of terabytes of data.

Mitigating Customer Risk Through Open Source

However, the decision to embrace open source was also strategically motivated by business considerations: utilizing open-source projects fosters customer confidence.

“A large multinational corporation will inevitably have a reliance on a startup,” Abernathey stated. “But by leveraging open-source tools, the customer’s risk is lessened, as they retain control of their data even if we were to change direction or cease operations.”

Notable Clients and Applications

Earthmover’s clientele includes insurance technology firm Kettle, which utilizes the platform for wildfire risk assessment, and RWE, a German multinational energy company.

Abernathey noted that companies in the renewable energy sector, vulnerable to weather fluctuations, employ Earthmover’s tools to predict supply and demand.

Democratizing Access to Weather and Geospatial Data

The overarching objective is to broaden access to weather and geospatial data. “For a trading desk, the ability to visualize the latest forecast on a dashboard is crucial,” Hamman explained. “They shouldn’t be required to execute a Python script to achieve this.”