Awais Ahmed and Kshitij Khandelwal are just 21. Not long ago they were reading about Nasa’s Voyager mission and Mars rovers, like most kids fascinated by space stories. Now they’re in Los Angeles visiting Nasa’s Jet Propulsion Laboratory (JPL), talking to scientists who worked on those Voyager and Mars missions about their startup, Pixxel.
Ahmed and Khandelwal co-founded Pixxel in their final year at BITS, Pilani, in 2018. This year, they raised pre-seed funding to start building microsatellites to beam down data for AI analytics. Next June, the first one will hitch a ride on an Isro rocket. They aim to send out two more next year, then six satellites every quarter in 2021 until a constellation of them can take an “ECG of the earth round the clock," as Ahmed puts it.
Speed of execution is the name of the game in entrepreneurship. A shot in the arm came from Pixxel being the only startup from Asia to be selected for the first batch of 10 startups in Techstars Starburst Space Accelerator in Los Angeles. Techstars is a leading accelerator and partners with corporations. Nasa’s JPL, US air force, Lockheed Martin, Maxar Technologies, and Israel Aerospace Industries are part of this programme. It gives the young entrepreneurs an early grasp of business use cases for their satellite as well as validation from the world’s top space scientists.
Fine-tuning Business Models
“JPL, Lockheed Martin and others are open to working with private entrepreneurs. It’s the same with the Israeli space industry. They’re always looking for a win-win," says Ahmed, who is also the CEO of Pixxel. “Our first adopters will be here in the US. So we’re getting inroads into the ecosystem and fine-tuning business models. Two years of reaching out to these people are condensed into three months till October (in the accelerator programme)."
But the high point for Ahmed was meeting seasoned Nasa scientists. “Getting their validation that we are on the right track means a lot," he says.
The idea came to them while working on an IBM Watson AI XPrize project, a $5-million competition for teams around the world to tackle “the world’s grand challenges," using IBM’s open-source AI platform.
The Bitsians wanted to build AI models to derive insights from satellite imagery to help farmers. Satellite images were available from open source repositories but they ran into a problem.
The images did not have the resolution their AI models needed. “We wanted to analyze the nutrient and moisture content of the soil, chlorophyll content of leaves, and so on," says Ahmed. “The available satellite image data could at best show crop health status, but little beyond that."
The same applies to other areas where satellite imagery can be transformative—such as in exploration for minerals, oil and gas. Just as infrared images provide night vision, light reflected from the Earth can help zero in on undiscovered mineral-rich areas to explore. But better images are needed for this. Top-grade defence satellites have the capability, but those are not available commercially and don’t cover all geographies.
“We reached out to organizations across the world. Then we decided to do it ourselves," says Ahmed. Instead of doing what they started out with—AI analytics on satellite imagery—they moved upstream to solve the more fundamental data problem.
Pixxel aims to be a web platform on the cloud. It will provide access to satellite images for AI analytics teams from various organizations and companies. Customers can specify the frequency—daily or weekly—and area of coverage.
It’s tempting for the startup to do the analytics too, but it doesn’t want to bite off more than it can chew. “Right now our focus is on the tougher part of the value chain. Whoever owns the data controls the entire supply chain," says Ahmed. “We will be a platform play, rather than provide customized solutions, because that will be more scalable."
Taking on Competition
There are a few startups in this space. The biggest of them is US-based Planet, which has 150 active satellites and $183 million in funding, according to Crunchbase. Its first set of 88 Dove satellites went into orbit aboard an Isro rocket in 2013.
But its images don’t have the resolution to go deeper than simple use cases, according to Ahmed. “Our instrument will provide more information," he says, without wanting to disclose more at this stage.
What’s hard is to strike a balance between the capability of the microsatellite and its weight, because every extra kilo can add ₹20 lakh to the launch cost. But doing both the hardware and the software to capture and process the image data has its pluses. “It gives us control over the pipeline to build the satellites according to the data required by customers," says Khandelwal, CTO of Pixxel.
The hardest challenge Pixxel has faced so far is not the technology, but raising funds for a capital-intensive venture as student entrepreneurs. “It took eight months. Everyone was interested but no one was willing to put their money where their mouth was. We had to be better than normal entrepreneurs because we were students, I think," says Ahmed.
Finally, the BITS Pilani alumni network kicked in. Bay Area’s Raju Reddy made the first bet. growX ventures became the first institutional investor in July this year. This will enable Pixxel to make its first launch, but it will have to raise more funding.
“Pixxel may have to look outside India for the next round of funding because VCs here are risk-averse," says Sheetal Bahl, partner at growX, an early- stage fund.
What could help Pixxel go forward is being part of the Techstars global network, if all goes well with their launch next year. It will be a giant leap for the two founders from their rural roots. Khandelwal is from Amrapur in the drought-prone Vidarbha region of Maharashtra, where his father has a farm and veterinary practice. Ahmed is from Aldur in Karnataka’s Chikmagaluru. His family has a pharmacy and a coffee plantation.
Ahmed went to Mangaluru for his pre-university studies before making it to BITS Pilani for MSc Mathematics. He became eligible for a BE Mechanical dual degree in the second year, but dropped it the following year because by then he was deep into more exciting extra-curricular activities.
He had joined a project called Ananth to work with Isro scientists to build nanosatellites in his first year. He founded Hyperloop India the following year to build a prototype of a pod that travels at the speed of sound inside a vacuum tube. It was the first team from India to present the prototype to SpaceX’s Elon Musk in the Hyperloop Pod competition, along with teams from MIT and Harvard.
It was the Ananth and Hyperloop experience and mentoring by Isro scientists that gave the Pixxel founders the confidence to build and launch satellites. “We stand on the shoulders of giants," says Ahmed.
*Sumit Chakraberty is a Contributing Editor with Mint