Investor Watch: Alphabet's Sidewalk IP is selling the ultimate upgrade.
Sometime in the 1960s or 70s, a Stanford University computer scientist coined a phrase that came to be known as ‘Amara’s Law’. He said that people tend to overestimate the impact of new technology in the short run but underestimate it in the long run.
Roy Amara was probably talking about computers, but he could just as easily have been talking about autonomous vehicles, first dreamt up in the 1930s and finally being considered for mass trials in China, Europe and North America from this year.
The current generation of road developers and asset management have underestimated this, says Jonathan Winer, co-CEO of Sidewalk Infrastructure Partners (SIP).
“Roads of the future will not be a passive asset – they will be disrupted by autonomous vehicles during the lifetime of many current PPP contracts,” says Winer. “A lot of people out there own legacy assets like toll roads and over the life of those projects there will be a disruption.”
SIP – spun out from Google-owner Alphabet last year after an abandoned a smart city project in Toronto – is the opposite of a passive infrastructure investor.
A lot of people out there own legacy assets like toll roads and over the life of those projects there will be a disruption
Winer’s team wants to use Cavnue – one of four ‘innovation platforms’ – to build the roads of the future, starting with a 42-mile test project in Michigan, currently at feasibility stage. Cavnue is a small start-up within a start-up, but it has big financial backers in the form of Canadian pension fund Ontario Teachers' Pension Plan (OTPP) and Alphabet, and it is moving at speed.
In February, the platform was chosen to work with road developer Transurban, Macquarie Capital and techtransit company Via, to ease congestion on a Maryland highway outside Washington DC and it is reported to be having conversations with around eight other governments for similar arrangements. The Michigan P3 project will complete first phase feasibility next year and is targeting operation two years later.
But SIP is not placing all its bets on self-driving cars.
As a “thematic investor”, says Winer, it plans to launch an additional two platforms this year, exploring advanced logistics, sustainable data centers and urban air mobility. And, in the process, it is disrupting the very model of infrastructure investment itself.
Fund disruptor
SIP has been structured as a private company, rather than an infrastructure fund, to take levels of technology risk rarely seen in the asset class.
It can take small stakes in technology start-ups and then invest long-term at asset-level to put this technology into wide-scale use, pricing risk whether it is venture-like or more infrastructure-like.
“Project finance doesn’t understand how to underwrite technology risk,” says Winer, “and raising additional venture money to fund these projects often doesn’t make sense because it is highly dilutive,” he says. “We follow an investment process which is slightly different than core or core-plus structured fund… with flexibility to invest early-seed stage to late-stage asset-level project financing [into] differentiated pipelines.”
SIP’s small team of finance veterans, data scientists and technologists raised USD 400m last year through Series A funding from OTPP and Alphabet.
Winer says about 80% of SIP’s balance sheet is to acquire and own infrastructure assets but declined to disclose how much has been invested to date. Each of the four innovation platforms has a target investment of around USD 100m but Cavnue is already expected to go far beyond that.
He says there is room for more investors to join the company as both asset owners and technology providers.
“We are not going to go out to 20 investors [but to] people with large USD 10bn plus infrastructure portfolios who are aware of the transformation and disruption coming from technology and want to be part of developing those new platforms and assets and be on the disruptive side of it,” says Winer. “Each partner can scale up their commitments quite significantly.”
Teaching technology
For now, SIP’s infrastructure partnership with OTPP is still scaling up.
The pension fund’s investment in SIP has been allocated from its ‘innovation platform’ known as TIP, which focuses on late-stage venture and growth equity, rather than its infrastructure fund.
Out of a five-member SIP board with several observer seats, OTPP has two representatives: Dale Burgess, who leads OTPP infrastructure; and Olivia Steedman, who leads TIP.
In 2019, Steedman explained the partnership in a statement on SIP: “We will bring our infrastructure and investment expertise to the table, while leveraging our partners’ world-leading capabilities in technology, to enable sustainable, intelligent and efficient physical infrastructure”.
At this table, Winer says there is now “active debate”, which helps OTPP to unlock asset management challenges and SIP to decide which platforms to create next.
“We have a list of the 20 big ideas for how technology could be changing infrastructure and in those meetings the Alphabet people will [give their view] and say ‘we are not quite there yet, we need to wait for the next generation’, and the Teachers guys are going ‘here is where our platform companies are feeling the pain – our airports are losing short-term parking, how do we offset that with technology?’"
“When we do focus, we dive deeply into the Alphabet ecosystem in terms of getting the right technology experts at the table and then we talk through the asset platforms that Teachers owns.”
It is not known how willing OTPP is to apply SIP’s technology to its infrastructure assets under ownership. Neither Steedman not Burgess responded to emails and an OTPP spokesperson declined to comment on the relationship with SIP.
According to Winer, SIP will typically “talk through the asset platforms that Teachers owns” to identity possible pilot sites – one of which was used for SIP’s waste recycling platform, but he adds: “There hasn’t been a material application [of our technology on OTPP assets] yet. There are several times when we have got close [such as] on the logistics side.”
Deployment risk
Where a typical closed-end infrastructure fund might face capital deployment risk, SIP must now identify ways to deploy the technology it owns.
As the number of platforms it owns increases, so does the requirement to apply this new technology across a greater range of sectors. For Cavnue, SIP has the support of Ford, GM and other major car manufacturers, and partners such as Transurban in Maryland, but it also needs access to operational infrastructure assets and pilot sites for testing.
This is not always easy.
In 2017, SIP affiliate Sidewalk Labs – which according to SIP's website backs SIP and previously employed Winer as head of investments – won a competition to build a ‘city of the future’ on small Toronto waterfront site.
According to reports, Sidewalk Labs unveiled a proposal over 190-acres which included sensors in traffic lights and garbage bins to monitor use, sidewalks that expand after rush hour and a thermal grid using sewers and factories for heating.
A public backlash, centred on data ownership and privacy, followed and the project was finally abandoned last year.
In the run-up to the cancellation, Jim Balsillie – one of Canada’s most known tech leaders and former CEO of Research In Motion (owner of BlackBerry) – described Sidewalk Labs’ Toronto plan as a “pseudo-tech dystopia” and “a colonizing experiment in surveillance capitalism”. Around the same time, a report from the National Research Council warned that Canada was in danger of becoming a “data cow” for foreign tech platforms.
We have a list of the 20 big ideas for how technology could be changing infrastructure
In February this year, another Sidewalk Labs smart city project – led by spin-off company Replica – was also ditched following data concerns. Other ambitious smart city projects globally have either failed to meet expectations or been scaled back, such as Masdar City in Abu Dhabi and Songdo near Seoul.
While the Toronto experiment did not involve SIP, it reveals the difficulty of deploying new technology into infrastructure at scale and of convincing public sector stakeholders and residents of its merit.
Not surprisingly, Winer is a techno-optimist but acknowledges not everyone might be.
He is also moving in increasingly political circles.
This month, SIP launched the Innovative Infrastructure Initiative (I3), a consortium of “technologists, investors, policymakers, academics” which counts a former director of the White House Office of Energy and Climate and a former US Deputy Secretary of Transportation as members.
“We haven’t seen the backlash as much, but it could happen,” says Winer, in reference to data and privacy concerns. For Cavnue’s Michigan road, the state’s department of transportation will own all the underlying data on road users, he adds.
“Municipals have gotten very sophisticated very quickly in understanding how to own and control their data. The fact that people are having that conversation [on data] is very encouraging. On the airports side, people are aware of the cameras and security everywhere [so] I don’t think there is an expectation of privacy, but the user experience is a very important one.”
As municipal authorities in North America face tax revenue shortfalls since the pandemic, the question of them working more closely with investors such as SIP has not gone away.
In Toronto, a team led by Alice Xu known as Connected Community (Smart City) – note the brackets – is now working on a digital infrastructure plan which Xu describes as “a blueprint of what a digital city should look like”.
“Cities are becoming more aware and trying to find creative and mutually beneficial ways to work with the ecosystem,” she says. “Developers are waiting for some oversight, or guidance come from the government.”
“I think digitization is not going to stop, it's not going to slow down when in fact it's only, I think, sped up. And so what people can do, and what I think the city government ought to do, is to find a way to place the people's voices and what people want at the center of the regulations.”
Virtual power
Users on another of SIP’s platforms may need less convincing on sharing their data.
Through its Resilia platform, SIP is collaborating with San Francisco-based OhmConnect to launch a virtual power plant that will incentivize customers to reduce their power usage when California’s outdated grid is at peak demand or close to blackout.
Founded in 2014, OhmConnect was expected to pay customers around USD 4m last year using the model.
Joining other demand response and distributed energy providers in the state such as Tesla, Leap, Enel X, CPower and Google Nest, SIP has invested USD 20m in OhmConnect directly and USD 80m in the new 550MW virtual power plant, which should add 750,000 customers to OhmConnect's 150,000 already active in California. OhmConnect also has operations in Australia and Canada and plans to expand to Texas this year.
SIP’s other two innovation platforms are CoFi and Polysift, which are focused on broadband and waste recycling respectively.
To create a new platform, SIP typically spends between three and six months developing a technology masterplan in consultation with Alphabet and experts outside the company. Each platform combines everything from incubations, intellectual property, growth equity and venture capital investments acquired from technology companies.
The city government ought to find a way to place the people's voices and what people want at the center of the regulations
So far the team has a hit rate of 50%: Winer says SIP has run this thematic process in eight areas and pushed four platforms over the line in a two-year period.
Crucially, SIP has been able to develop these platforms without the pressure to generate returns in year one or year two, as faced by most closed-end infrastructure funds. Along the way, it hopes to create a pipeline of investible projects in which to deploy its technology.
“If you look at how much time and energy goes into in the first couple of years developing these kinds of innovation platforms, you really want to be able to follow-on and build-out large, differentiated pipelines of either assets you are acquiring and platforming, or greenfield assets,” says Winer.
By the time it finally has an operational infrastructure asset of its own, Cavnue will have travelled many more years than most traditional road P3s.
According to Winer, the equity on the project alone is significantly more than the USD 100m target for each SIP platform, without including the investment in the technology company itself. During the Michigan P3’s feasibility, Cavnue must also identify a financing structure for it and other roads where it is partnering with developers.
Winer sees a potential revenue uplift by incrementally charging users more for vehicles with advanced driving systems or for the reduction in time or cost of getting into the lane. But technology can also increase the enforcement potential for users who take advantage of high occupancy toll (HOT) lanes to “more actively manage the road,” he says.
In this way, the P3 is more like a hybrid of a bus rapid transit lane and a HOT lane, he says, but the model is also being considered as a replacement for planned light rail transit routes which may now face lower traffic projections since COVID-19 upended commuter demand.
SIP’s disruptive approach to infrastructure is not limited to sectors such as roads and power grids, or to infrastructure investment, but may soon apply to contracting and procurement models such as P3s, says Winer.
There is a movement towards more innovation-based P3s in the US, he says, offering the example of Maryland which in a recent P3 RFP asked to see a strategy for both shared mobility transit and advanced driving systems and autonomous vehicles from bidders. Most developers don’t have that suite of technology, says Winer.
“We are entering into a world, and it is already happening with some procurements, in which you will no longer be able to bid unless your road management system isn’t state of the art.”
“Infrastructure as an asset class is going through a foundational change. Digitalisation is transforming how infrastructure assets are designed, how they are implemented, how they are operated and as a result how contracting works, how offtake and PPPs work, fundamentally changing the risk-return profile of these assets.”
The new risk profile on some of these assets is becoming clear, but it will be some years until the returns are known. Let’s hope Roy Amara’s law doesn’t apply to how they are being estimated.