India-based space-tech startup KaleidEO plans to launch four edge computing-enabled satellites in 2026, offering unique capabilities compared to existing EO constellations. The launch initially slated for 2025 has now been revised to a newer timeline because of course correction.
“Considering our current position, we have decided to revise our timeline, shifting our target from 2025 to 2026. This adjustment is necessary because the current state of technology, our partnerships, and the overall integration of these elements require further investment in technology development and process refinement,” explained Arpan Sahoo, the co-founder and COO of KaleidEO, in a recent interview with AIM.
Sahoo said that the objective was to achieve a similar performance with fewer satellites, rather than launching 10 to 20 of them. This led them to step back and rethink the design of their payload.
By innovating on the swath width, the company aims to cover the globe with just four satellites, as opposed to the 12-14 typically required by commercial constellations. “What we are rethinking is let’s not make the same mistake,” Sahoo remarked, referring to the commercial viability issues faced by some companies in the past.
“At the heart of it, what it really does is makes your data and your space infrastructure commercially viable,” he emphasised.
The prototype payload, tested aboard an aircraft over the UK and Austria, captured images in 5 spectral bands (red, blue, green, near-infrared, red-edge) at a spatial resolution of 16 cm. However, the satellite version set to launch in 2026 will capture images globally at 1m resolution with a 65 km swath in six bands.
“It gives us about six bands, which are red, green, blue, infrared, red, edge and panchromatic. So panchromatic is like a culmination, on average, of all the bands,” he said. This balance allows KaleidEO to target multiple use cases, including agriculture, infrastructure monitoring, forestry, and climate impact analysis.
Edge Computing FTW
The company is also emphasising on the use of edge computing capabilities, which it has achieved to make EO data commercially viable and efficient.
These capabilities enable satellites to run deep learning algorithms and provide near real-time insights, with up to a 99% reduction in data volume. Additionally, optical inter-satellite links can increase data throughput from 500 Mbps to 5-10 Gbps, significantly improving data transfer speeds.
In a similar effort, the company has partnered with Finnish startup ReOrbit to launch a software-defined satellite with ESA in Q2 2025. As part of it, the company is combining ReOrbit’s software with its own AI models to make Earth observation data more accessible.
This would enable satellites to network in orbit and create intelligent, interconnected satellites for efficient data delivery. “ReOrbit is working with ESA on a program called InCubed,” explained Sahoo. “The idea about the entirety of this program is to demonstrate commercial viability and scalability for commercial users, not just government or strategic users,” he added.
KaleidEO’s Role
KaleidEO’s role in the partnership is to integrate various systems, including ground stations, cloud servers, and analytics solutions, to provide an end-to-end EO infrastructure as a service.
“KaleidEO serves as the unifying force, integrating all these systems into a cohesive whole. Meanwhile, SatSure contributes its comprehensive solution stack, and ReOrbit takes the lead as the primary asset provider,” Sahoo explained.
The company has already onboarded several clients, with one expressing interest in utilising an optical ground link between the satellite and a direct optical station on the ground.
According to Sahoo, this technology could potentially increase data throughput from 500 megabits per second to an impressive 5-10 gigabits per second, significantly reducing power requirements and improving efficiency.
Overcoming Technical Challenges in Edge Computing
The COO also highlighted the challenges of running deep learning algorithms on satellite imagery in real-time, likening it to the processing power required for gaming GPUs. “Your GPU is an edge device. On the ground, we can pump in 250 to 500 watts of power into a GPU card and extract as much benefit or performance out of it,” he explained.
“The primary limitation in space arises from the severe power constraints. Instead of having access to 250 watts, you are restricted to a mere 10 or 15 watts, as you rely solely on solar energy. In space, every resource is precious and limited,” he elaborated.
To overcome these limitations, KaleidEO has employed supercomputing algorithms and hardware-level programming, resulting in a remarkable 99% reduction in processing time. “What would normally take 15 minutes, we are doing in 10 seconds today,” Sahoo stated.
Success Stories and Future Prospects
Meanwhile, Sahoo also shared with AIM a success story involving the Airport Authority of India, highlighting the impact of KaleidEO’s technology. “SatSure, KaleidEO’s parent company did a grant project with them about two-and-a-half-years ago. It was kind of intuitive. You would want to know what are the obstacles in your approach and take off vectors,” he explained.
Using satellite imagery and advanced modelling techniques, SatSure was able to identify obstacles in the buffer areas surrounding airports, ensuring safer operations.
“Today, India is probably the fifth country in the world which has access to this technology at an international level. And there are so many countries with far older aviation industries. But we are the fifth now in terms of having technologies where digital innovation through space has happened recently,” Sahoo proudly stated.
As KaleidEO continues to expand its offerings with edge computing and submeter imaging capabilities, the company has expressed excitement about exploring new use cases and market segments.
Its commitment to innovation and ability to deliver end-to-end solutions has already attracted a diverse range of clients, including Reliance General Insurance and the Airport Authority of India. With its cutting-edge technology and customer-centric approach, KaleidEO is well-positioned to drive the adoption of EO infrastructure across various industries.
Enabling New Use Cases and Market Segments
KaleidEO’s edge computing and submeter imaging capabilities are expected to open up new use cases and market segments. Sahoo emphasised the importance of integrating hardware and software to push the boundaries of EO infrastructure.
“Our focus lies in achieving a more seamless integration of hardware and software components. We are pushing the boundaries of what EO infrastructure represents today and redefine how EO infrastructure as a service should look in the future,” he stated.
SatSure has already made significant strides in the agriculture sector, enabling more than 2 million farmer loans over the past five years through its technology partnerships with banks. The company has also worked with government organisations to digitise farmland and consolidate data for subsidy and resource management.
EO Boom
The earth observation (EO) data industry in India is experiencing significant growth, with an expected CAGR of 15.34% during the forecast period from 2024 to 2029. On the other hand, on a global scale it is expected to reach $4.43 billion by 2025, up from $2.74 billion in 2019 and is projected to grow at a CAGR of 6.52% to reach $5.54 billion by 2029.
The growth is driven by emerging use cases across various sectors, including urban planning, waste management, carbon emissions tracking, and vegetation monitoring.
Technological advancements like edge computing at KeleidEO are also contributing to the growth of the EO data industry with companies launching constellations of small satellites in LEO to provide more frequent and high-resolution imagery.
