Musk, Mittal’s one small step in space is one giant leap of faith

If sky's the limit, trust mavericks to reach for outer space! Elon Musk is one such dreampreneur who wants to transport humans in rockets between continents and to colonise Mars, and, hence, referred to as a ‘futurist’ and (unkindly) ‘insane’. While some of his plans are plain audacious, others seem realistic.

One of his pet projects is to provide internet connectivity through a mega constellation of miniature satellites, which will float in the lower earth orbit (LEO). Interestingly, joining Musk in the LEO-satellite race is India’s leading telecom czar Sunil Mittal. While Mittal runs the world’s second-largest terrestrial mobile service through Bharti Airtel, he has now set sights on space communication, albeit in his personal capacity. This Musk-Mittal’s internet-from-space odyssey seems fantastical, but it warrants a dekko because they are planning this for the second-most populated country on planet earth.

Space tech has come a long way from 1945 when Arthur Clarke believed that just three satellites could meet the world's growing communication needs. Instead, we are now living with 2,666 satellites, of which 1,918 are in the LEO, as per the Union of Concerned Scientists, the US-based science advocacy group.

LEO occupies the space at an altitude between 500 and 1,500 km above the earth’s surface and is commonly used for data communication, remote sensing, and human space flight. But since one satellite in the LEO takes 88 to 127 minutes to orbit around the earth once, a string or constellation of satellites are needed to ensure uninterrupted connectivity (See: In a different orbit). “LEO satellites can’t run individually, they run as the mega-constellations. If you don’t have a mega constellation, you cannot say, I’ll give you broadband for 10 minutes, and then you wait 40 minutes, till another satellite orbits around. It has to be available all the time. It has to be seamless just like we enjoy roaming with mobile tech,” explains Shivaji Chatterjee, SVP and head- enterprise business, Hughes Communications India.

Geostationary satellites can cover a larger area—of a third of the globe against a mere 1/66th (1.51%) by a LEO satellite. But since the LEO offers high bandwidth and low communication time lag, commonly known as latency, new age space players such as SpaceX and OneWeb with better satellite technologies, economies of scale, and a glut of capital are leaning towards these satellites.

The early history of LEO satellite connectivity is not exactly inspiring, given the spectacular failure of Motorola’s Iridium in 2000. After a $5-billion investment, the venture had to be finally bailed out by the US government and, today, Iridium operates with a limited number of satellites offering communication services largely to the maritime industry. Similarly, Teledesic, too, had attracted investors such as Bill Gates, Saudi Prince, Motorola, and Boeing, but wound up in 2002.

Bill Wiseman, senior partner, McKinsey & Company, says that all but Iridium scaled back or cancelled their intended constellations because of high costs and limited demand. “All suffered financial problems. After that experience, many industry analysts and investors remain sceptical about the viability of large LEO constellations,” says Wiseman.

Today hope has been rekindled since the demand for connectivity has increased manifold. “A combination of different market demand, new business models, technology progress, and availability of capital,” says Wiseman. Indeed. Getting into orbit has become comparatively less expensive, even as satellites and their manufacturing methods have become more advanced —for instance, Iridium’s satellites weigh 850 kg vs 50 kg for the current generation of LEO satellites.

Closer home, Mansi Kedia, policy researcher, the ‎Indian Council for Research on International Economic Relations too, believes this time it will be different. “It might be game-changing. Mobile (usage) took off because so many private players backed the technology. The same will play out in satellite broadband. It will not be a peripheral technology anymore and might become mainstream for a particular segment of broadband users,” she says.

That is exciting and, if the buzz among private players is anything to go by, they are feeling it too.

Coming of age
Like always, with cutting-edge technology, the US remains the hotbed of LEO launches too. The Federal Communications Commission’s (FCC) roster list features the likes of Starlink, OneWeb and Amazon, besides Telesat, Keppler, and Leosat, which are looking to deploy thousands of small satellites. There is also talk of Apple and Facebook joining the race. But for now, it's Musk and Mittal who are at the forefront of the new wave.

As of March 2021, SpaceX has in space 1,100 satellites out of its target of 12,000 satellites. It has even kicked off beta services in rural parts of the US. Starlink, as the internet service is known, will have satellites with four-phased-array antennas, while the end-user terminal or receiver features a dish on a motor that will automatically steer towards an optimal degree of the satellite’s antennas. Though not official, Reddit community users have claimed that Starlink is delivering download speeds of 80-150 Mbps and upload speeds of about 30 Mbps. Musk is eventually, promising download speeds of 1 Gbps to 10 Gbps but only after 800 satellites are operational.

To put this into perspective, the fastest internet connection in India now gives an average 4G download speed of 15.4 Mbps and average upload speeds of 7.2 Mbps.

Compared to SpaceX, OneWeb has a much smaller fleet of 110 LEO satellites transiting in space and a target of 648 such satellites (See: Jostling for space). Shravin Mittal, managing director, Bharti Global, told Outlook Business, “Growing data consumption, requirements of low latency, and the lowest cost to build make OneWeb a compelling proposition for global connectivity. OneWeb will provide high-speed connectivity and complement terrestrial networks.” Complementing terrestrial is for a reason.

Sunil Mittal in an interview with Satellite Today said, “It is almost impossible to cover the last billion people with fibre or radio terrestrial networks. It is too expensive and too time-consuming and for the past 25 years, I and my entire telecoms fraternity have tried and tried, but have not been able to go to the last unconnected people.” 

This is Mittal’s second coming in OneWeb. He had earlier acquired a stake in the company in 2015 along with a consortium comprising Virgin Group, Qualcomm, Hughes and Airbus. Mittal exited the company later by selling his stake to Softbank. Referring to the exit, Mittal told Satellite Today that it was because “there were too many pushes and pulls from different shareholders” and “couldn’t see the company moving in the direction I wanted.” However, OneWeb, after raising over $3 billion in capital and sending an initial batch of satellites into space, filed for bankruptcy past year.

That’s when, Bharti Global, the family investment vehicle of Sunil Mittal run by his younger son Shravin, won a bid for 45% along with the UK government (45%) to rescue the company from bankruptcy. On why he chose to back the company, Mittal has said, “LEO constellations can cost anywhere from $5.5 billion to $7 billion. If such a start-up was presented to me now, I would not take it. But there was a large amount of money already spent. So, with the $2.5 billion we are proposing to spend, this will be the cheapest constellation anybody has ever had. As new shareholders, we get the benefits of the past investment, this makes economic sense.”

What also worked as a sweetener was that 74 satellites were already launched before Chapter 11 (the bankruptcy process), and this meant OneWeb had delivered on the mandated 10% fleet requirement of the International Telecommunications Union and had secured its priority rights to operate in Ku-band spectrum. “That made it attractive,” revealed Mittal.

But unlike Starlink, which owns the entire value chain from making satellites to launches, OneWeb is based in the UK, its satellites are made in the US and they are lofted into orbit by a Russian space vehicle. However, Mittal is more than excited with the deal. With a base station network comprising of OneWeb’s 650, 220,000 in India and 40,000 in Africa, Mittal believes he now “covers the whole globe, every inch of water, every inch of the forest. Why should this not be exciting for a telecom company? Everything else on top of this including enterprise, aviation, maritime and defence contracts, are additional revenue opportunities.” Chatterjee believes the go-to market will be a little different for Mittal, “While Bharati will look for partners outside Africa and Asia for OneWeb, within Africa and Asia, probably use it as a part of their own service delivery for cellular backhaul.”

Even as services are yet to be commercialised, Musk and Mittal are already making plans for India. “We remain on track to becoming the world’s first global LEO constellation, with 24x7 coverage over 50 degrees North starting November this year and have full coverage by June 2022,” says Shravin Mittal. While OneWeb hopes to start services in 2022, an aggressive Starlink has already opened bookings in India. More on that later.

The single biggest trigger for entering India is that traditional telcos, despite growing mobile subscriber base, find delivering data connectivity in rural areas and hinterlands a challenge.

India calling
It’s been over two decades since India embraced the internet revolution. From 10 million internet users in 2,000, we are now 726 million; largely post the rollout of 3G and 4G mobile services. Though the Narendra Modi government has plans for a Digital India, the big challenge is that broadband penetration in rural India is limited to 29 subscribers per 100. Given that the government wants to e-link all its welfare and subsidy programmes, there is a dire need for uninterrupted internet connectivity in these regions. According to reports, Bihar, Uttar Pradesh, Jammu and Kashmir, Madhya Pradesh and West Bengal have the least number of internet subscribers with an average of 32 per 100. A report by Crisil says that only 25-30% of the country is connected through the fibre network, despite the advent of 4G and this number has to go up to 70% if the country has to deploy 5G.

The telecom backbone that knits the country ends abruptly at the towns and larger villages. Beyond that, cellular coverage extends only up to a radius of 5 km. “Telecom companies may need investments of up to around  Rs.1 trillion only in laying fibre networks over the next two to three years. Higher land cost and right-of-way approvals make fiberisation cost per km as high as nearly  Rs.10 million per kilometre in metros,” mentions Hetal Gandhi, director at Crisil Research. Most telecom players deploy microwave to connect towers as laying fibre is cumbersome, expensive and tangled up in legal wrangles. Given that telecoms are already saddled with a staggering debt of about  Rs.4.3 trillion, in addition to  Rs.1.4 trillion adjusted-gross-revenue (AGR) dues to the government, they can’t spend much. They did try experimenting with proprietary broadband technologies such as WiMAX-D, but the attempts came to nothing owing to high cost and low volumes.

In fact, K Ramchand, member (technology) at the Department of Telecommunications (DoT), was candid in his remarks at an event of the Broadband India Forum, when he said, “It’s not possible, seeing the terrain and geography of this country to reach all these villages through fibre and terrestrial micro-wave communications. There will be areas where they need to get connected through satellite communications.” The Telecom Regulatory Authority of India (Trai), which wants to increase the minimum broadband speed from 512 kbps to 2 mbps, has sought comments on a consultation paper titled “Roadmap to Promote Broadband Connectivity and Enhanced Broadband Speed” issued in August 2020. Trai’s agenda envisions fixed-line broadband access to 50% of households by 2022, universal broadband connectivity at 50 Mbps and 100 Mbps broadband on-demand to all key development institutions, along with 1 Gbps connectivity to all gram panchayats by 2020 and 10 Gbps by 2022.

This ambition cannot be met through Indian Space Research Organisation (ISRO) alone. The official agency for satcom in India has been plagued by capacity shortages and rents a large part of the capacity from overseas private satellite firms. This deficit will hurt more as data consumption rises— data consumption in India has risen 100x between March 2015 and March 2020, from 0.1 Gb per subscriber per month to 11 Gb.

Not surprising that, in its feedback to Trai’s paper, SpaceX has asked the telecom regulator to pave the way for satellite broadband. Patricia Cooper, vice president, Satellite Government Affairs at SpaceX, in the letter to Trai has stated, “A diverse mix of broadband technology platforms and service providers, including those new to India’s market, will best accelerate broadband access across the nation.”

Cooper did not respond to an email sent by Outlook Business seeking comments on the article.

In the letter, SpaceX has also highlighted how its solution would eliminate the high-costs associated with traditional wired broadband, “SpaceX does not require expensive ‘last-mile’ fibre lines in order to deliver reliable high-speed broadband. In fact, the ‘last-mile’ for SpaceX’s Starlink satellite system consists of the Ku-band connection from the consumer’s home directly to a satellite in orbit, entirely eliminating the largest cost inhibitor to near-term universal broadband coverage in India.” Besides, SpaceX in its six-point agenda has also asked for blanket licensing tools and a band-splitting model to encourage sharing of spectrum between private telecom operators in India.

Though OneWeb has not commented on Trai’s paper, Mittal in his interview stated, “We are very excited about India’s broadband program, which has been trying to get to the last village. ISRO chairman and we were connected on day one, even before we acquired this.”

In response to a Trai question on why fixed broadband speeds were so low in India, SpaceX said that policies discouraging the participation of high-speed satellite internet technologies such as Starlink’s in broadband deployment played a role in driving up costs of fixed-line broadband and making speeds slower. “Terrestrial fixed wired and wireless technologies have brought economic and social advancement across India, but come with an inherent infrastructure expense based on a cost per km that is difficult to scale while maintaining affordable prices for end-users when connecting remote or rural communities,” SpaceX said, adding that satellite broadband would reduce cost-per-km in rural areas.

The approaches taken by Musk and Mittal to address these gaps are different though.

Split wide open
While Musk wants to disrupt telcos by going directly to consumers, Mittal is looking at a partnership-based approach. Hoping that the satellite broadband will be green-lighted, Starlink has already opened bookings in India. It has sought a one-time fee of $99 ( Rs.7,250). “Availability is limited. Orders will be fulfilled on a first-come, first-served basis,” reads the announcement on the Starlink website. Chatterjee feels the exercise is more about creating an excitement around the service and also raising easy money. “I mean, that’s a good way of crowdsourcing. But I really feel that doesn’t tell the whole story,” says Chatterjee.

The Reddit community for Starlink, which has been tracking the system’s rollout, has mentioned that Starlink has expanded to select areas in over 30 US states and seven provinces in Canada. The company has also taken the service overseas to the UK, Germany and New Zealand. Starlink has revealed that it is targeting coverage in several areas in India by 2022. Users have to punch in their location to know whether the service is available and, for now, reports say that users can book from Maharashtra, Gujarat and Madhya Pradesh.

Chatterjee believes the direct to consumer strategy has always been Musk’s approach. “He has modelled his business on the lines of Apple, which has a direct connection with consumers. Starlink is trying to do the same thing where he’s saying, I will acquire customers directly, I will go to the market, I have a super brand.”

In complete contrast to Musk, Mittal wants to take the partnership approach with other telcos. “I am not in competition with them. SpaceX wants to be in competition with them. They want to go directly to the customer. I am not going to make this mistake, because in my simple assessment, wherever a terrestrial network is present, there is very little opportunity for satellites. I am where they are not,” said Mittal in the interview.

McKinsey’s Wiseman believes that these are still early days to comment on whose strategy will work well. “We think it is too early in the game to predict how different business models will fare; we expect a few years of experimentation,” he says.

Final word on the strategy will be decided by the costs involved and what works in India.

Still expensive
Musk has estimated that his string of pearls in space will cost him more than $10 billion, though the cost of launching each batch of 60 satellites falls by about $1m per launch. The big expense would be towards the Starship system, the reusable heavy-lift launch vehicle under development by SpaceX and estimated to cost $2 billion to $3 billion. Meanwhile, building rocket prototypes will continue to cost $200 million each, four times more expensive than a Falcon 9 rocket, notes the website Motley Fool.

For now, SpaceX has raised $850 million in fresh equity in a new round this March, pushing its valuation to $74 billion, up 60% over its previous round in August 2020. Analysts at investment bank Cowen in a report have estimated that Starlink will only have enough capacity to serve 485,000 users even if it has 12,000 satellites in orbit. Therefore, it will need copious amounts of capital to keep its launch pads firing and the latest funding doesn’t change SpaceX’s long-term plan to spin off Starlink as a public company.

OneWeb has managed to raise $1 billion in new funding— $500 million each from the UK government and Bharti Global. The company had raised a much higher amount of $3.4 billion before it ran into financial difficulty with SoftBank, its largest backer, refusing to invest additional capital. Initially, the target for OneWeb was to offer low-cost internet links for individuals, schools and small businesses across developing regions, but now the focus is on providing connectivity to entities able to pay higher access fees, such as airlines, cruise operators and governments.

Musk has claimed that Starlink’s end-user terminals are easy to set up– plugin socket, point at sky – but the cost of these terminals remains the biggest unsolved challenge. The reason is that the antenna on the dish is complex and costs around $1,000, but Starlink takes only a one-time $499 fee for its satellite dish and Wi-Fi router. SpaceX has reportedly said that mass-market production could bring the cost down to $300. Vivekanand Subbaraman, analyst at Ambit Capital, is sceptical of the pricing. “Telcos in the US charge $70 a month. Why would you sell something which is so cheap,” he asks.

Musk wants the end-user terminals to be running for more than five years. “You can’t send people to service these things because a lot of these places will be in the middle of nowhere. So the fully considered cost of the user terminal is the hardest thing for Starlink or any space-based system for the general public,” Musk was quoted as saying.

Musk’s concerns on getting the cost right is also an outcome of the fact that investors backing Starlink would be eager to see the project take off. Concurring with the view, Wiseman of McKinsey says “We don’t believe in infinite patience or capital [from investors], so at some point the companies will need to show that the business case closes. We think that continuing to bring costs down is critical.”

Against this backdrop, pricing the satcom broadband service will not be easy in India.

Cost sensitive
In India, the current data tariff is  Rs.300 ($4.13) a month for 1 GB a day and Starlink’s offer is for $99 a month in the US. The latter’s offer is nearly 23x times more expensive than the average mobile-internet plan in India. The multiple could go higher if the average tariff in any hinterland is taken into account.

Ashok Jhunjhunwala, professor of electrical engineering in IIT Madras and who has served as director on boards of various telcos, believes satellite broadband is not going to be mainstream in India. “In the previous decade, the average plan cost around  Rs.120-130 which, after the advent of 4G, has come to around a little over  Rs. 200. That speaks volumes about how price sensitive the Indian market is,” says Jhunjhunwala, who had authored a research paper titled “Connecting Rural India with Broadband Wireless.”

Highlighting the challenges in reaching 3G connectivity to outer regions, Jhunjhunwala points out that over the past decade telcos have managed to bridge the gap with 4G technology. “(But) the government has to make a clear choice, on whether it wants money or wants good connectivity in the hinterlands. By taxing telcos at 33%, it has made it tough for telcos to invest in improving their reach. If they can lower the tax burden by 10%, and with 5G technology, telcos can address these connectivity challenges,” he says.

Concurring with Jhunjhunwala, Subbaraman of Ambit Capital says, “With the presence of high grade commercial network which offer data at such low prices, the likelihood of something like this succeeding and grabbing large part of consumer need for the internet seems highly implausible.”

While India is price sensitive, it is also where expensive cars and premium services are lapped up in high numbers. Take the case of Netflix, which has 203 million users globally of which India accounts for 2.4 million users, as per data from Comparitech. So, Starlink may happily play the Netflix of broadband internet, even as it advocates the need for improving connectivity in remote corners. Currently, at the beta-testing level, Starlink has 10,000 users in the US and analysts estimate that it would need three million subscribers in three years to break even.

Given that Musk has already talked about satellite technology struggling to supply enough internet bandwidth to users residing in heavily populated locations, reaching critical mass would be a challenge for Starlink. “Starlink is great for low to medium population density. But satellites are not great for high-density urban areas. So you’re actually better off having 5G for that,” Musk told popular podcaster Joe Rogan. BK Syngal, chairman of TEMA 6G Council and former chairman of BSNL, too agrees. “It will not be a ubiquitous service provider, but be a gap-filler where broadband services don’t exist,” he says.

Policy researcher Kedia, however, believes different business models can be built around satellite broadband to make it more affordable in Tier-II cities and the hinterland. “While satellite broadband will have good demand in Tier-I and II cities, where private mobile players aren’t aggressively investing, in rural areas we could see a second level of distribution model emerge, in which a connection is purchased by some institution at the gram panchayat level and is resold to users,” says Kedia. This could be possible given that the government has initiated the Bharat Net project, a broadband project aimed at connecting 250,000 gram panchayats, by using funds from the Universal Service Obligation Fund (USOF). Syngal too believes the government can subsidise the service by using the corpus of the USOF. For instance, in the US, SpaceX is enlisted to receive $886 million over 10 years under the Federal Communication Commission’s Rural Digital Opportunity Fund.

Wiseman feels that, in the end, it will all boil down to economics and physics. “Some places (whether because of terrain or remoteness) are unlikely to ever offer good returns for installing terrestrial infrastructure. In the long run, the question is whether that offers enough return for the satcom providers and, if not, then how much business they can capture in areas where they do need to compete with terrestrial providers?”

But an answer to that won’t be clear as Indian’s telecom journey has been bumpy with policy hiccups and challenges and it’s unlikely to be a smooth beaming for satellite broadband in India.

Seeking right signals
While the government has made its digital ambition clear, a lot depends on whether it will be able to put in place a comprehensive communication policy that ensures smooth rollout for satellite broadband in India and, more importantly, happily co-exist with telcos. Given that India's communication satellites launches by ISRO since 2000 has been slow to take off (See: Slow Space) there is a growing need for private sector participation.

To begin with, the ISRO’s draft Spacecom Policy guidelines bar foreign companies from introducing satellite-based broadband in India. This means Starlink’s plans to beam into India would need a new legislative and policy framework, as Cooper mentions in the letter. ISRO has created a new commercial-arm NewSpace India that will carry R&D work of the space agency, co-produce PSLV and launch satellites. “We have to see if the government wants to retain its monopoly or not let go as they did in telecom, where BSNL exists in whatever form, and it’s the private sector that is driving the growth in the market. ISRO has a set narrative of getting Atmanirbhar and Make in India without realising that, by the time we become self-reliant, space tech would have progressed 10-fold. It’s better to leverage on what we are good at, rather than trying to do everything ourselves,” feels Kedia.

Among its other comments, SpaceX has also sought protection for satellite service in high-frequency bands which include a set of spectrum identified for 5G services. Cooper’s letter reads, “...while India has long encouraged satellite operators to deploy gateway earth station facilities within the country, this policy is thwarted by the absence of Ka-band (26.5-40 Ghz band) frequency assignments that are required to communicate with those gateway earth stations.” The company has also asked for extending the existing licencing regime for multiple, identical satellite terminals, often referred to as a “blanket licence”. Such a “blanket licence” would streamline the site-to-site licencing requirements and accelerate widespread deployment of two-way satellite broadband terminals to support innovative satellite services in India.

While OneWeb has secured global priority rights in the 6 GHz spectrum and hopes to start services in 2021 across the northern hemisphere, India could feature on the map sometime mid-2022; Bharti will still need to wait for the new space legislation to come in place.

Mittal sounded confident in the interview quoted earlier, “We have had tremendous support from the Department of Space and ISRO. We were talking to them about the low-cost manufacture of user terminals and even in the first constellation of satellites. The good thing is that the prime minister is driving the conversation in India.”

Currently, the Space Activities Bill is with the Prime Minister’s Office and will go for inter-ministerial consultation before being brought before Parliament for approval. Given that PM Modi had in 2017 visited Tesla’s HQ and met Elon Musk, an accommodative policy for private participation could be expected. Kedia says, “I feel if Musk and Mittal steer the narrative a lot better, then we might see a shift in government’s focus in terms of loosening reins on availability of spectrum, which is one of the biggest challenges, and regulations around participation of foreign companies.”

Given that Modi has made a strong privatisation pitch recently with his remark, “Can babus do everything?”, it’s more a question of when and not whether private players can enter the India space.

The milky way
If both Mittal and Musk get the green signal, their strategy in India will be similar to what they are doing in other markets. Bharti already runs an enterprise business that offers VSAT services and Mittal will use OneWeb to cater to the same segment and not threaten to disrupt his own profitable cellular business in the country. OneWeb, for instance, offers infra services for two of its first customers, Talia and Intermatica. Mittal himself is clear of where the opportunity lies when he mentions in the interview, "...cellular backhaul, rural broadband, enterprise and customers like Amazon that have depots and factories in large rural areas. Defence and police forces, and a large maritime play in the Indian Ocean.”

Similarly, OneWeb could also target the emerging area of automation for the space industry, where BSNL entered into a partnership with start-up Skylo to deploy a satellite-based narrow band-internet of things network, targeted at industries such as fishing, agriculture, construction, mining and logistics. The solution by Skylo will connect with BSNL’s satellite ground infrastructure and provide coverage across India, including the seas. “We are convinced there are numerous use cases for this capability including but not limited to maritime, aviation, IoT, cellular backhaul,” says Shravin Mittal, “OneWeb has made significant progress over the last quarter, with the latest launch scheduled for March 25, 2021.” However, Chatterjee feels that given the nature of the technology, timelines cannot get extended. “Because of the nature of these new systems, you cannot go live partially, that's the worst publicity to get. So, unless your constellation is up to 80% level, you cannot give a 24-hour service anywhere in the world.”

Despite scepticism, the battle lines have already been drawn in space with Mittal not thinking too highly of his competition. “Stripping- out networks in favour of a disruptor space-delivered strategy makes them [SpaceX] at best a competitor of note and, at worst, exemplars of predatory, below market pricing.”

Brave words and now we need to see who wins in the battle—the futurist-realist or the realist-futurist.