SpaceX’s Transporter-18 mission carried TakeMe2Space’s MOI-1A satellite, EON Space Labs’ MIRA imaging system, and SatLeo Labs’ TAPAS-1 thermal payload.
Processing imagery in orbit can reduce the volume of data sent to Earth, while thermal and multispectral imaging can support more specialised monitoring services.
TakeMe2Space has signed up 23 customers for its in-orbit computing services and plans to complete a six-satellite constellation by 2028.
Indian spacetech startups TakeMe2Space, SatLeo Labs, and EON Space Labs sent AI computing and Earth observation technologies into orbit aboard SpaceX’s Falcon 9 today.
The Transporter-18 rideshare mission lifted off from California, carrying TakeMe2Space’s MOI-1A satellite with EON Space Labs’ MIRA imaging system onboard. SatLeo Labs’ TAPAS-1 thermal-imaging payload flew on Dhruva Space’s LEAP-2 satellite.
Hyderabad-based TakeMe2Space’s MOI-1A is designed to process Earth observation data in orbit using AI, allowing customers to receive insights without downloading large volumes of raw imagery.
Ahmedabad-based SatLeo’s TAPAS-1 is designed to capture land-surface temperature data, while Hyderabad-based EON’s MIRA captures imagery across multiple spectral bands.
MOI-1A is a 14-kg satellite with onboard computing capacity of 117 trillion operations per second (TOPS), 2 TB of storage, and a nine-band multispectral imager.
Customers can deploy AI models directly on the satellite to analyse imagery in orbit, instead of downloading it to a ground station for processing. This can reduce the volume of data that needs to be downloaded by a factor of 1,000, TakeMe2Space founder and CEO Ronak Kumar Samantray told Inc42. The satellite identifies objects or patterns in imagery and sends back the results.
He explained that a 2,000-km stretch of synthetic-aperture radar (SAR) imagery could generate around 300 GB of data. If an AI model is instead used to identify specific objects or patterns in orbit, the resulting output could shrink to only a few kilobytes.
The startup’s first-generation satellites are designed around inference rather than AI model training.
TakeMe2Space has signed up 23 customers for its in-orbit inference services. Around 80% are outside India, according to Samantray. These include international geospatial companies, while Indian educational institutions are using the infrastructure as an orbital research laboratory.
The use cases currently span agriculture, defence, supply chain intelligence, mining and insurance.
MOI-1A follows TakeMe2Space’s 2024 technology demonstrator, which validated onboard AI inference and data processing. The startup subsequently lost MOI-1 during the unsuccessful PSLV-C62 mission earlier this year.
Samantray said the broad satellite architecture remains similar, but individual subsystems have undergone multiple development cycles since the earlier mission.
The startup expects to make MOI-1A commercially available shortly after reaching orbit. It plans to launch its other Gen-1 satellite, MOI-1B, in November aboard ISRO’s SSLV. That satellite will carry a black-and-white imager primarily intended for object detection.
TakeMe2Space then plans to launch larger spacecraft as it builds a six-satellite constellation.
Its Gen-2 satellites will have roughly 10X the capacity of the Gen-1 spacecraft, according to Samantray, with around 100 TB of storage per satellite and significantly greater computing capacity.
The startup also plans to introduce laser-based links between satellites, allowing them to exchange data without routing it through ground stations. For instance, one satellite could pass processed data to another with a better connection to a ground station for transmission to Earth.
Two Gen-2 satellites, MOI-2 and MOI-3, are planned for launch in 2027, most likely aboard Indian launch vehicles. Samantray said the startup has not booked these launches yet because he expects Indian launch services to operate more regularly by then.
The startup has secured a slot aboard a SpaceX Falcon 9 mission in late 2028 for MOI-4 and MOI-5, which would complete the constellation, he added.
Samantray expects each Gen-2 satellite to generate around $1 Mn in annual revenue. The startup is targeting about $6 Mn in annual recurring revenue from its in-orbit computing business by 2028.
It is also designing Gen-3 satellites in the 1,500-kg class with around 25 kW of power, with a potential launch around 2029, he said.
SatLeo Labs’ TAPAS-1, which the startup describes as India’s first dedicated commercial thermal payload, uses thermal infrared sensing to capture temperature information. Potential applications include urban heat mapping, crop-stress monitoring, disaster monitoring, and infrastructure surveillance.
The payload flew aboard Dhruva Space’s LEAP-2 satellite, which also carries a star tracker from France’s Sodern. The mission will test TAPAS-1’s sensing, calibration, and processing capabilities in orbit.
EON Space Labs’ MIRA, meanwhile, is a compact multispectral imaging system weighing 502 grams. Integrated with MOI-1A, it is designed to capture Earth observation imagery for applications spanning agriculture, mining, marine monitoring, supply chains, and insurance.
Together, the missions bring specialised imaging and onboard data processing into focus as India’s spacetech startups work towards commercialising their technologies.
IN-SPACe chairman Pawan Goenka called the launch a “big day for India’s private space sector”, saying it showed the “range and ambition” of India’s private space ecosystem.
Source: Inc42
