At Antarikchya Pratisthan Nepal (Space Academy Nepal), Trishna Shrestha and her senior team of satellite engineers are rolling up their sleeves as they prepare for the launch of Munal in a few months.
Munal is a 1U (10x10x10cm) CubeSat weighing 1–1.3kg. For the first time, it is designed by a team of young students and engineering graduates in Nepal. This project is set to mark a momentous occasion in the satellite history of the nation as it is viewed as an advanced and improved version of the first satellite, NepaliSat-1, which was previously developed under the Birds-3 Project of the Japanese Kyushu Institute of Technology in 2019.
NepaliSat-1 was a milestone, paving the way for space research and innovation in Nepal. And now, Munal furthers this progress. It is Nepal’s first high school satellite equipped with what could be the first step to an imaging mission with an inbuilt AI. The success of this project would mean there is a bigger possibility of scaling up the size or number of satellites in the future. As the satellite is planned to be launched by ISRO (Indian Space Research Organisation), the experience of beginning a collaboration with India will hopefully open up the opportunity to use their facilities for future projects.
‘Munal is advanced in such respects as we are using two cameras for remote sensing and better data keeping and processing,’ said Trishna Shrestha, Satellite Research Fellow at Antarikchya Pratisthan Nepal. Unlike the previous satellite, which only had a single camera for taking RGB (red, green, blue) images, Munal has an additional camera for IR (infrared) imaging. RGB images are all coloured images, while IR images focus on texture rather than colour. These two images are combined together for NDVI (Normal Difference Vegetation Index) which is a remote sensing technique used to assess the health and density of vegetation.
Munal is the first time AI has been incorporated in this way. With the CubeSat, the research organisation is trying to collect AI-generated data and planning to study vegetation patterns in Nepal. Once launched, Munal will be orbiting 400km from Earth. Two inbuilt cameras used for this mission will take the pictures from space.
To improve remote sensing, an AI algorithm has been set in the camera, which conducts an image segmentation test to make the collected images and data more accurate. The AI model is installed in the controller of these two cameras. It is used to classify images into either good or bad, cloudy or non-cloudy, earth, space, sunlight, or moon. After the image is classified, the image segmentation process is used to determine what portion of the image occupies which part. ‘This way we will have better information on the image before we download it […] which will save us time,’ Shrestha said.
‘Munal also reflects the symbol of historical and cultural representation of Nepal through audio-visual elements incorporated in the CubeSat,’ said Rishav Adhikari, a Satellite Research fellow at APN. ‘This is like a Voyager spin on Nepal,’ he quipped.
CubeSats generally have a lifeline of two years, and the data shall be collected until it de-orbits. For the operation, APN has to build a ground station to communicate with Munal so that they can download all the data collected in space. For this, they will need to design a specific antenna, and once it is set, the team can send a command from the ground station to the satellite in space and can download the required data or image.
‘For now, we are only trying to collect the data to analyse what portion of Nepal has a healthy vegetation pattern. Through this, we can collect information on barren lands. If this mission is successful, we can modify and update [the satellite] further to get details on disaster-prone regions and make an early warning system in the future,’ Shrestha added.
‘The innovation journey since the inception of APN has been daunting but we have just landed at the beginning of our long-term vision,’ said Abhas Maskey, founder of Antarikchya Pratisthan Nepal. ‘At APN, we are trying to break the glass ceiling by challenging our horizons and paving the way for bigger innovation. Munal is special in that respect because it marks the beginning of making a Nepali satellite constellation in space. For our long-term goal, we are in the process of building a system to collect real-time data,’ Maskey explained.
Lack of space policy, component procurement, and fund delays are the existing challenges that have impacted the timeline of project completion. However, Maskey feels a sense of support sprouting at the government level.
Despite existing gaps, APN has been conducting tinkering training programs on satellite space systems and has so far trained over 1,500 school students across the country. Maskey says the major aim of the tinkering programs is to fuel the curious minds of school students at the local level so that they can mitigate the human resource gap in future projects. ‘The learning space and mentoring that APN has provided has aided my growth as a fresh engineering graduate,’ said Anjali Dhami, a member of staff training at APN who is currently working on a different project in collaboration with Kathmandu Metropolitan City. With an aim to put a Nepali astronaut in space by 2050, APN is trying to push its boundaries more than ever. ‘We are working with international space organisations like MBRSC (Mohammed Bin Rashid Space Centre) located in Dubai and making a payload for their satellite. The next plan is to collaborate with bigger space organisations so that we can grow better and live up to our word, spirit, and objective,’ Shrestha remarked.