Investment Notes: Esper Satellites

For the past 20 months, we've been keeping a close eye on Esper, watching as they steadily made progress. We were excited when the team, led by Shoaib and Joey, shared the details of this exciting and fast-paced fundraise. In these investment notes, I will dive into two things:  

1. how Shoaib Iqbal and Przemyslaw ‘Joey’ Lorenczak are advancing hyperspectral imaging technology for enhanced Earth Observation with Esper Satellites; and

2. how we got conviction on what the Esper is building through Club Investible and Advisory Network—a vast global network of experts that shaped and validated our investment thesis  

Beyond Multispectral: How Hyperspectral Imaging is transforming remote sensing

Imagine looking at the world through a lens that reveals secrets hidden beyond human vision. That’s exactly what multispectral imagery (MSI) does. This decades-old technology captures images across multiple wavelengths of the electromagnetic spectrum, far beyond just what our eyes can perceive.

Think of it like this—while we see the world in the spectrum of RBG, MSI extends that view to include non-visible wavelengths like near-infrared (NIR) and shortwave infrared (SWIR). These extra layers of information allow scientists, farmers, and environmentalists to uncover details that would otherwise remain invisible—whether it’s assessing crop health, monitoring deforestation, or analysing mineral compositions in the ground.

MSI sensors work by capturing data in a limited number of broad spectral bands, making them efficient and well-suited for applications where massive amounts of data aren’t necessary. Whether it's tracking changes in land cover or studying environmental shifts, MSI has been a powerful tool for decades, offering a deeper understanding of our world, one wavelength at a time.

Nearly two decades after MSI made its mark in spectral technological advancement, hyperspectral imaging (HSI) pushed the boundaries of what MSI could do as it captures continuous spectral data across hundreds of narrow bands. It was first explored in airborne remote sensing missions for military and scientific applications.  

The surge in computation power increased the potential of HSI, as HSI sensors could now analyse materials with an unmatched level of precision, detecting subtle differences in composition, health and characteristics of objects and terrain across earth. Just like a fingerprint, each material reflects light in a unique way, and HSI is able to capture those distinct spectral signatures—allowing end users to differentiate between materials, identify substances and uncover details invisible to the naked eye. Initially confined to defence and high-end research due to its steep costs and complex processing demands, HSI is now breaking into the commercial world, revolutionising precision agriculture, industrial quality control and optimising resource extraction for mining companies. What was once an exclusive tool for intelligence and research is now shaping industries and helping in their decarbonisation plans.  

A USD 3.8 Trillion opportunity up for grabs in the Earth Observation market

A study by WEF with Deloitte— Amplifying the Global Value of Earth Observation—estimated that the potential value add from Earth data is estimated to reach USD 700B in 2030 with a cumulative USD 3.8T contribution to global GDP between 2023-2030.

As we look at the landscape of Earth Observation (EO), it’s clear that its potential goes far beyond what we see on the surface. In fact, EO has the power to inform interventions that could reduce GHG emissions by over 2 billion gigatons of CO2 equivalent each year—a game-changer in our fight against climate change.

Most of EO's value isn’t just in the data itself, but in how it's applied across industries. The value each sector can unlock from Earth data depends on its technology readiness, global reach, and its ability to scale in the future.

Industries that are already tech-savvy are likely to tap into the power of EO data the quickest, unlocking immense value in the process. However, it’s the more traditional industries—the ones that still have a long way to go in adapting—that stand to gain the most in the long run. The potential for EO to disrupt and revolutionize these sectors is staggering, making it clear that this technology has a bright future ahead.

High reward for extremely high costs: barriers to hyperspectral imaging widespread adoption

While hyperspectral imaging offers unparalleled insights across multiple industries, there are multiple obstacles preventing that from happening, such as the high sensor costs, limiting lack of a common standard for manufacturing of hyperspectral sensors, insufficient labelled data for training, and the high volume of produced data.

1. High costs of satellites and hyperspectral technologies: Hyperspectral sensors and cameras are expensive to manufacture, as these sensors are highly sophisticated and often custom-built. These tend to be large and power intensive, to power the data processing infrastructure required onboard, which poses challenges in fitting and integrating into satellite systems.  

2. High volume of produced data: Collecting hyperspectral data face challenges in storing and transmitting the enormous file sizes inherent when gathering such high-dimensional data with hundreds of spectral bands. This data must be stored onboard the satellite and transmitted to the ground from the satellite, which can be expensive. Thus, in addition to the challenges of gathering useful data, the volume of data also needs to be taken into consideration.  

3. Processing complexity and data interpretation: Hyperspectral data requires advanced models capable of handling large datasets and performing advanced analysis techniques such like dimensionality reduction, spectral feature extraction and classification algorithms to interpret the rich spectral data within the image. HSIs often requires significant post-processing time, limiting its real-time processing capabilities.  

Esper Satellites is creating a “fingerprint scanner from space”, delivered at a fraction of the price

Esper is challenging the status quo by delivering high quality hyperspectral images at a fraction of the costs and is further reducing the costs by building a network of satellites with their advanced hyperspectral smart sensors. But how do they do it?  

Compactification of Imager Design

While the key ingredient to the recipe remains a chef’s trade secret, Esper have been successful in developing a compact imager design.  

Since we first came across them in May 2023, Esper has launched three satellites into space. In 2024 alone, Esper had launched ‘Esperesso’, their imaging sensor demonstrator in 3 missions. With ‘Esperesso’, Esper had miniaturised a hyperspectral design into a CubeSat form factor, which reduces the cost per wavelength by 800%.  

Esper has also successfully space qualified and integrated high-quality components into their sensors, allowing them to provide accurate and detailed atmospheric and surface level chemical data of the Earth. Their sensors are able to capture multiple bands at high spatial resolution, while their onboard software decodes the chemical data collected from the sensors. In their latest two satellite demonstration launches, they successfully reduced launch costs by 20 times compared to their competitors.  

While still heavier than the industry standard at 12 U (vs 3U industry standard), ‘Esperesso’ outperforms industry benchmarks in capturing a spectral range of 100 bands between 400-1000nm at a lower cost per wavelength.

Piggybacking on larger satellites for zero latency capacity

Esper, despite being a small payload, leverages the full data throughput (i.e. communication capacity) of a much larger host satellite SpaceX’s Transporter-10 mission. Typically, transmitting such large datasets from space faces bandwidth constraints, but since Esper operates using the throughput of a much larger satellite, it can potentially overcome these limitations.

By adopting this space architecture, Esper avoids the need for its own dedicated infrastructure while still benefiting from high-speed data transmission.

This setup allows Esper to function as if it were a high-performance satellite while remaining a small, cost-effective payload, enhancing real-time space data capabilities.

In-house spectrum derivation algorithm with hyper parallelisation capabilities

Esper has developed an optical system designed to capture raw light and project it onto their sensors. To make sense of this data, they’ve developed an algorithm that processes and optimizes the information, onboard.

At the core of this is Spectrum Derivation, a statistical method that refines spectral bands based on lab calibration data. This allows them to take raw light and break it down into hundreds of precise bands, extracting useful data from each wavelength.

Much like how software-defined radios adjust frequencies through software, their approach lets us define spectral accuracy and bandwidth programmatically. This flexibility allows Esper to create customized data products based on specific user needs, making their system both adaptable and efficient.

In the works, Esper is building a hyper-fast compute module designed to process its algorithm directly onboard its satellites—a breakthrough that will unlock real-time insights from space like never before. Partnering with Spiral Blue, Esper leverages cutting-edge space-based computers to integrate its SD algorithm with modern computing power, creating a key competitive advantage. By overcoming processing bottlenecks, Esper is making hyperspectral data more accessible and actionable than ever before.

The Investible investment process is backed by a global network of domain and industry experts

At Investible, we know that it takes a village to do investments and portfolio management. That’s our not-so-secret sauce as to how we are able to move fast in our due diligence process to make 17 new investments and 14 follow-on investments in 2024.  

We often turn to Club Investible members, our Climate Tech Advisory Board, and the founders in our portfolio—experts in their respective fields—to gain deeper insights into both technology and the market. Their firsthand experience helps us understand real-world challenges and opportunities.

One key challenge that stood out in agriculture is the difficulty of accessing reliable hyperspectral imagery from satellite suppliers. Farmers and agritech companies need high-quality data, but securing it consistently has been a major hurdle. That’s where Esper comes in—working to become a dependable source of high-quality hyperspectral imagery tailored for agricultural needs.

Another valuable insight was the market’s shift from custom, boutique solutions to mass-manufactured, scalable technology. This shift aligns perfectly with Esper’s approach, positioning it to meet the growing demand for widely accessible, high-quality hyperspectral imaging at scale. As the industry evolves, Esper is right at the intersection of need and opportunity, ready to bridge the gap.

Our global advisors are a key source of guidance, helping us navigate challenges and stay ahead of industry trends. Their insights are invaluable in shaping informed decisions. In our work with Esper, their detailed feedback refined our investment thesis, highlighting key risks and how Esper intends to addresse them. Their expertise is essential, giving us the confidence to move forward with clarity.

Launching five additional satellites…and beyond!

Esper is on track to launch their fourth satellite ‘Esperesso’ later this year, followed by its first flagship imaging payload ‘Esperoko’ in 2026 and another three satellites thereafter. We’ve had a glimpse into Esper’s product roadmap, and it’s an exciting vision coming to life. Their new satellites are targeted to accelerate mineral exploration and extraction while minimizing fossil fuel impact and ensuring regulatory compliance. We’re thrilled to be part of their journey as we watch Esper turn this ambitious goal of building a network of satellites and bringing it to orbit!

‍