Over the next decade, ride-hailing platforms like Uber, Lyft, and Didi are now expected transition from human-driven, internal combustion engine cars to autonomous electric vehicles – including millions of driverless cars, and potentially thousands of pilotless vertical take-off and landing aircraft.

By some estimates, shared autonomous vehicles won't dominate ground transport until 2050. Yet if we assume a more aggressive adoption S-curve, similar to that of smart phones over the past decade, up to 95% of U.S. passenger miles travelled could be served by 2030, by on-demand autonomous electric vehicles owned by fleets, not individuals.

The 2030 timeline assumes widespread regulatory approvals in the early 2020s, and powerful economic incentives for key stakeholders.

For fleet operators, autonomous electric vehicles could provide up to ten times higher vehicle-utilization rates and substantially lower maintenance, energy, finance and insurance costs. For consumers, this could translate to about half the all-in cost per mile for transport in 2030 that car owners pay today, not to mention an 80% reduction in accident rates.

According to a study of downtown Boston conducted in 2017 by the World Economic Forum, BCG, and the MIT Media Lab, a faster shift from privately owned vehicles to on-demand, autonomous electric vehicle fleets would result in 28% reduction of cars on the road, a 30% reduction in average travel times, a 66% reduction in CO2 emissions, and 48% reduction in parking spaces required.

It should be noted, that an aggressive transition to self-driving vehicles is by no means a given, nor will progress be uniform across regions, countries, or even cities. For such a transition to occur, multi-tier political and regulatory support and stability will be required. This will determine, to a large extent, the investment available for the provision of key infrastructural elements, including ubiquitous vehicle charging, and 5G connectivity to provide ultra-reliable, high-bandwidth wireless communications between vehicles, transport infrastructure, and passengers.

The road to 5G

5G will initially exist as an overlay to existing 4G networks, and will start achieving meaningful scale and reach by the early 2020s. 5G network operators will need to create new data-driven revenue streams.

The key challenge to scaling up the supporting 5G infrastructure is that it will be expensive to deploy and require fresh thinking on financing, business models, and regulation. Who will pay for 5G infrastructure, how broad will the infrastructure’s reach be, can it be commonly owned, and is there scope for effective public-private partnerships in the build out and operations?

Autonomous transport providers will certainly look for ways to optimize a blend of on-board processing and off-board guidance to ensure some level of working autonomy not fully dependent on cellular network connectivity. Nonetheless, the very AI powering driverless systems will improve with the consumption of ever larger volumes of live data generated by vehicles, infrastructure, and passengers.

By making driving, mapping, and video data available in real-time, 5G will extend the ability of vehicles to perceive their environment beyond line of site, through robust vehicle to vehicle and vehicle to infrastructure connectivity to coordinate to improve traffic safety, efficiency, and flow.

The rise of the 'passenger economy'

A report from Intel predicts the rise of a "passenger economy" worth $800 billion in 2035 as people in cars have time on their hands to consume goods and services, from immersive entertainment to doctors' appointments.

At least initially, autonomous transport fleet operators would be likely to establish command centers, where in the event of unfamiliar or challenging road conditions, a human operator could track cars on the road, click on one, and teleport into what the car sees.

Who will deploy and operate 5G infrastructure?

5G networks will likely be deployed by mobile operators and equipment vendors, leveraging software-centric, open source architecture, and built upon a mix of public and private physical infrastructure assets.

IHS estimates 19.6% of the total calculated 5G sales enablement of $12.3 trillion in 2035 will be to the automotive/transport ecosystem. This translates to more than $2.4 trillion in economic output across the automotive, transportation and logistics sectors and their customers.

Building out the necessary infrastructure will require significant time and investment to densify radio access networks and connect high capacity “small cells” to fibre. This is because 5G connections will use higher frequency radio waves that cannot travel as far as low frequency 4G and 3G waves.

What regulatory models will attract investment in the 5G services?

5G network operators will require flexibility from regulators to enter into spectrum sharing and infrastructure sharing agreements. There will be a need to streamline local siting/rights of way approvals for the installation of small cells and optical fiber that will interconnect them, including reasonable provision of access to public properties on reasonable terms.

The critical safety considerations of autonomous transport requires resilience by design in the planning of supporting network communications infrastructure, traffic management systems, and of course, the vehicles themselves.

Considering the volumes of potentially sensitive data that might be generated by autonomous transport systems, the broader ecosystem and relevant local authorities must agree which actors own which sort of data and which must act as data stewards; clearly defining the responsibilities for each.

These stakeholders may also explore the creation of data commons to incentivize the sharing of data. This could allow, for examples, local transport authorities to leverage the data generated to optimize road usage with differential pricing based on local traffic conditions, prompting passengers to optimize mode of transport, travel route and time of travel.

What will be the implications for accessibility?

Sharing economy companies have faced several challenges entering cities everywhere, and have often been challenged by the existing economic models of cities. Trust remains a key bottleneck issue. A new survey from the National League of Cities finds 33% of U.S. cities have a “very poor” relationship with sharing companies, but more than half said the relationships were “very good” or “good,” and 7% chose the word “tenuous.”

In cases where 5G coverage gaps emerge (for example lower income, or less densely populated areas), how might national and/or local authorities suitably incentivize network operators to deploy and operate the necessary infrastructure assets?

Because of high initial deployment costs, 5G is likely to be focused in densely populated, high income areas with particularly high density of network traffic such as downtown business and shopping districts, airports, train stations, stadiums.

Local authorities will need to ensure equitable, universal access to low cost autonomous transport as a service. This will only be possible with substantial investment to expand access to seamless vehicle charging and communication infrastructure.

Similar network economics for existing taxi and ride-hailing services show that drivers tend to gravitate to those areas where they can pick up more passengers and earn more revenues and return on investment. Policy planners will need to deal with a specific challenge from network economics. How broad, geographically, should the next generation transport, energy, and communication infrastructure reach?

Can public-private partnerships play a role accelerating the deployment and extending the reach of 5G infrastructure?

Economic leaders are currently pondering what kind of co-ownership models of 5G infrastructure – and of the data available from that infrastructure – can be imagined. They wonder if for example shared-ownership models of the infrastructure are possible.

Some ideas from past infrastructure financing models could be imagined to allow public-private actors to share in the upside of 5G investments. For example, Special Purpose Vehicle (SPV)s, could be launched to allow participation in 5G contracts, and allow for flexible financial structuring, allowing equity participation of private sector, public sector, and globally crowd-funded token sales.

The public sector might contribute spectrum, public properties, and tax waivers in exchange for equity/tokens in SPV.

Private sector companies would have a task in deploying their expertise and record in operating and maintaining infrastructure assets. As it is still early days, the co-ownership models of critical assets will begin to be explored. PPPs would remain important in this space because with 5G data, as we have discussed, is an economically critical asset, yet in turn while common ownership of it may be desirable, it certainly remains a challenge.