Flying cars and hyperloops?

Despite the promises made by governments to transform the world of car travel, 2050 probably will not look that different from today at the global level and that goes for many other aspects of travel beyond just cars. One clear obstacle to change is that vehicles have a long useful life. Though commercial fleet owners tend to replace vehicles after five years to avoid rising repair costs, private owners will still buy them and keep them running for far longer.

Total distance travelled is a better indicator of a car’s lifespan than its chronological age. Unless it’s destined to become a collector’s trophy, the chances are it will do more than 100,000 miles before being deemed too clapped-out to carry on. Although scrappage bounties do have an effect, in some countries the average age has steadily been creeping up over the past decade. Belgium reported the average age increasing from 14 to 17 years from 2012 to 2020. So, there is a good chance that petrol cars sold a few years ahead of the EU’s 2035 deadline for sales will still be on the road at the middle of the century.

Much hinges on how well governments decide to promote replacements for petrol-driven vehicles. According to the International Energy Agency (IEA), worldwide subsidies for electric vehicles (EVs) doubled in 2021 to almost $30bn. Seeing the signs, vehicle makers have been ploughing money into EV development themselves with investment in R&D and spending that at least matches the current subsidies. Using the projections from corporations’ annual reports, Reuters estimated last year that automobile makers around the world would spend more than $500bn on researching and making EVs and batteries by the end of 2030, with one-fifth of that spend coming from Volkswagen alone.

Russia’s invasion of Ukraine, however, has put a dent in the economic argument for owning an EV in the short term. As fears of shortages pushed oil and natural gas prices higher, the cost charging EVs rose by more than 40 per cent in just four months, almost eliminating the advantage on per-mile basis they have over petrol vehicles. A survey of UK drivers in the summer by the Automobile Association found almost two-thirds had been put off the idea of switching to an EV by rising energy costs.

Though the rise in charging prices seems likely to be a short-term effect, at least in more developed economies, it demonstrates how sensitive the overall market is to changes in relative cost, though buyers are not entirely coldly rational in how they assess the difference. When they put together their estimates of how total ownership cost would affect EV purchases, a team from the Massachusetts Institute of Technology’s (MIT’s) Energy Initiative gave the upfront purchase price double the weighting that a purely economic analysis would on the basis that car buyers do focus more on that element. This is not good news for EVs.

The IEA calculated that EVs in Europe and the US worked out as much as 50 per cent more expensive than petrol-driven equivalents in 2021. As the drivetrain is so much simpler than that of a petrol-driven car, which in turn means lower service costs, much of that difference is down to the battery packs, which account for at least a third of the sticker price. According to MIT, the battery packs are getting cheaper but the pace of reductions is slowing and, without a large change in chemistry, may bottom out above the crucial $100/kWh level. Bloomberg New Energy Finance (BloombergNEF) said the 2021 price came down to $132/kWh but would likely rise again in 2022 because the raw materials became more expensive.

The high up-front cost of electric vehicles will be a larger drag on their adoption where wages are lower. Analysts from BloombergNEF expect a two-tier market to continue though the coming decade in which emerging markets account for a fraction of EV sales compared to developed markets, only beginning to catch up in the decade to 2040. Even then, while 80 per cent of new car sales in the leading markets will be of EVs, developing markets may only scrape past 50 per cent by what is expected to be a sales peak in 2040.

To get zero emissions by 2050, sales of EVs have to rise faster; according to BloombergNEF, almost double the analysts’ current estimates for the period between 2035 and 2045 for what they see as the most likely scenario based primarily on expected economic factors.

One way in which change could come faster is if cars learn to drive themselves more reliably than humans. But optimism on how quickly that happens varies widely. We are already further from the first public demonstration of a self-driving car than from the, admittedly arbitrary, target year of 2050. After seven years of work on the project, in October 1994 Ernst Dickmanns, a researcher at the Bundeswehr University in Munich, packed dozens of Inmos transputers fed with images from several cameras into two Mercedes S500 cars and had them drive, with human supervision, along the motorways around Charles de Gaulle airport in Paris.

Edwin Olson, professor of electrical engineering at the University of Michigan and co-founder of May Mobility, argues that the key metric for predicting the success of self-driving vehicles is “miles per disengagement”: that is, how often a human is forced to take control of a situation to avoid an accident. His target is 10 billion miles, which is about 50 per cent more than the total miles driven worldwide in 2017. Olson’s most recent prediction is that, assuming vehicle makers continue improving ADAS (advanced driver assistance systems) software at the same rate they have done since the mid-2000’s DARPA challenges where robot cars occasionally mistook rocks for other cars, those lines will not intersect until around 2035.

One way of bringing that date forward is to have autonomous vehicles only work in certain situations, such as convoys on a motorway, or for fairly simple, low-speed urban driving as some do now where they crawl forward in a traffic jam but hand control back to the driver once the congestion clears. May Mobility itself focuses on sub-25mph driving on routes programmed into the ADAS. But it will likely be some years before we stop thinking of there being a specific seat for the driver. Some vehicles may resemble a lounge on wheels by 2050, but most of the others on the road will have a human at the wheel even if they are not fully engaged in the activity.

One of the most striking aspects of the Covid lockdown was how clearly one could hear the birds singing in cities without the background noise of traffic. The promise of the mobility revolution is that the peace could be restored when vehicles move to electric – albeit with some artificial sound effects to stop them being too quiet – and city dwellers give up on owning cars in favour of hailing a robotaxi whenever they feel like it. Will they? And will fleets of cheap robotaxis make cities less congested? The prognosis is not good on either count, according to work by the Energy Initiative at the MIT.

In 2019, the MIT group published a report that argued some assumptions about changes in attitudes to car ownership are overplayed. The theory is that younger generations do not see the car as a status symbol in the same way as the baby boomers and will be more willing to rent vehicles to get around. However, the study found that once you allow for factors such as disposable and urban living, ‘car pride’ remains as strong as ever. A potentially bigger problem for global emissions, given that even with electric vehicles powered by renewable energy there will be an environmental cost in manufacturing cars that mostly sit on drives, is that less mature car markets have higher levels of car pride. France and Japan are unusual in being less keen on owning cars than much of the rest of the world.

The younger generations in cities do not tend to buy vehicles, not least because parking is so problematic. But they will as they move into the suburbs, the MIT group reckons. And they will only be too keen to use robotaxis if they become available at the right price. This could lead to cities becoming more congested because the people using those short-term rentals will substitute for trips on mass public transport systems rather than selling a car to avoid its fixed costs. This may mean governments have to put in place policies that promote integrated transport, using pricing to make a trip to a station and using a train or bus for part of the journey cheaper than a longer-distance trip from door to door in the robotaxi. However, delivering such an integrated system may be easier said than done in a privatised system where different modes of transport will be competing with each other.

If you want to know what shipping will look like in 2050, take a look at shipbuilders in the Far East. As the average lifespan of a vessel is anywhere between 30 and 50 years, the chances are that whatever is waiting to launch  right now will be going strong when the middle of the century arrives. That does not mean shipping is not due for some major changes. It is even possible that changes in legislation mean some of those ships being built today will wind up with a different power source within a couple of decades.

According to shipping insurer DNV, companies having ships built for them today are seriously looking at ordering designs that will run on liquefied natural gas at launch but which can be converted a decade or more from now to run on ammonia. Right now, it looks to be a more likely replacement for fossil fuels than the other option of hydrogen – even though some of the tankers may wind up conveying hydrogen in some form from Africa or Chile to Europe and manufacturing centres along the Pacific coast of Asia.

Work has started on developing engines that can run on ammonia, which can be made using renewables, using green hydrogen as an intermediate step. MAN Energy Solutions expects the first of its two-stroke ammonia-fuelled engines to go into production in 2024, assuming trials go well. But there is also the issue of convincing shippers to switch fuels.  

One option for kickstarting the use of ammonia as a maritime fuel is to adapt the tankers that carry the gas to plants that make fertiliser: they can simply divert some of the cargo to the fuel lines. Ammonia happens to be a better carrier for hydrogen itself, which guarantees it a place in any future economy that relies on green hydrogen.

Why is ammonia better? It is far easier to liquefy than hydrogen. Less than 1 per cent of the energy is lost to liquefaction, compared to as much as 40 per cent for hydrogen. It does not escape from containers as readily as hydrogen and its energy density is better; more than 30 per cent higher than the far lighter gas. Unfortunately, it has issues. It burns more slowly than any other candidate fuel and engines will need to be carefully designed and maintained to avoid spewing out nitrous oxide as a waste gas – an even worse greenhouse gas than carbon dioxide – though this problem also faces hydrogen if burnt rather than passed through a fuel cell because the heat oxidises elemental nitrogen in air.

Over the past decade, the software content of cars has soared to the point that even manufacturers are beginning to think of them as data centres on wheels. They aim to manage their contents the same way: replacing software whenever the owner wants a new feature or when updates are needed.

Two trends are pushing carmakers towards the idea of the software-defined vehicle and away from traditional families of models with features fixed at the time of sale. One is electrification.

As the powertrain gets simpler, with electrically powered motors attached directly to the wheels in place of a complex and intricate systems of gears and crankshafts running through the spine of the vehicle, it opens up the possibility of treating the entire car design as a set of interchangeable modules. Several manufacturers have already chosen to opt for a single frame on top of which they can deploy different bodies, battery pack combinations and computer hardware.

The second driving force is that of creeping autonomy, in which manufacturers gradually augment the ability of the car to drive itself, taking advantage of advances in machine learning and sensor processing to do so. The manufacturers also perceive software updates as a way of selling more aftermarket services, such as headlights that dip in response to oncoming traffic and smarter cruise control through to novelty sound effects for electric vehicles and, most puzzling of all, a monthly subscription to be able to activate a built-in seatwarmer.

One possible outcome though not yet on the roadmap of most manufacturers is whether hardware modularity will be their way to build more sustainable products. In principle, if they obtain more money through rented services, there is less incentive for them to push full replacement of old models. Instead, they may simply have older computers, seats, motors and battery packs swapped out as they wear out. However, as private vehicles are already comparatively long-lived, modularity may not be an advantage if the battery, motors and frame all have similar useful lifetimes.

Fleet owners may opt to gradually upgrade and adapt their vehicles instead of selling them into the private-owner markets after a few years as they do now. Launched in 2021, the EU’s Urbanized project is an attempt to look at the viability of this approach for light commercial vehicles and whether the use of easily swappable body parts will cut the number of vehicles companies need to buy to handle seasonal differences in operations.

One trend that Covid has helped push along is the drawn-out death of the traditional high street in favour of next-day delivery and even supermarkets beginning to resemble online-delivery warehouses, with shoppers and staff competing for what’s on the shelves.

Naturally, as online deliveries have become more common, the number of miles being clocked up by last-mile deliveries has soared over the past decade. Like mobility as a service, this is contributing to an increase in the number of miles travelled across the country by light delivery vehicles (LDVs).

Within cities, councils have been encouraging couriers to move away from relying on LDVs. In 2020, the City of London decided it wanted to set up a network of van-like rickshaws to carry parcels around its environs and deploy small-scale logistics centres across five of its car parks to support delivery by foot in the Square Mile by 2025. Amazon has begun deliveries around the nearby borough of Hackney using a similar scheme. At the other end of the technology scale, the Isle of Wight and Milton Keynes have worked with logistics companies to trial various kinds of robots: airborne drones on the island during the Covid pandemic to handle NHS deliveries, and wheeled robots to deliver groceries and takeaway food in Milton Keynes.

Though electrically powered robots will be cleaner than diesel LDVs and become more so as renewables continue to build out, the key to fewer delivery miles may lie less in the last-mile technology but in the logistics management. Several studies have proposed more work on the City of London’s approach of creating smaller satellite hubs, even making them mobile to adapt to daily or hourly changes in demand. This would, in principle, allow for more deliveries to be aggregated. Customers may be encouraged through discounts to accept slower delivery and have them aggregated to cut the overall number of trips made by each courier. As with the integrated transport needed to make mobility as a service less energy intensive, the key will be to find ways to make the suppliers and logistics operators work more closely together, which may prove to be more of a stumbling block than technology.

One way self-driving vehicles or even just those with advanced safety computers will improve on what humans can do at the wheel lies in their ability to communicate beyond flashing indicator lights and making hand signals of varying levels of politeness.

Standards are already in place for vehicles to send messages to each other not only as they pass but, by relaying them through the cellular network or transceivers attached to lights and signs, when they are out of sight. This will give them the ability to detect queues of traffic and other roadblocks long before they come into sight, so that cars can slow down early.

There is another potential benefit from vehicle-to-everything (V2X) communications to drivers, though not necessarily all of them at the same time. The same signals can inform traffic-management systems of the best time to change the lights at a junction or direct vehicles down one road in favour of another. One trial conducted by Ford earlier this year upgraded eight traffic lights in Aachen, Germany with radio units to detect approaching ambulances so they could change in favour of the emergency vehicle as it approached. The city of Detroit in the US has made a similar upgrade to favour snowploughs in winter and several projects are assessing whether it makes sense to give priority to heavy goods vehicles where possible to save on the energy they consume when speeding back up.

A key obstacle to V2X, compared to just vehicle-to-vehicle communications, is one of cost. The Aachen trial involved equipping traffic signals with the same kind of LiDAR, cameras and computers as those used in vehicles with driver assistance on top of the cellular transceiver. A study by engineering group Ricardo that focused on adding 5G coverage to street furniture found that it would, at a minimum, cost in the realms of €500 per kilometre, largely because of the cost of attaching the additional radio units to the core network, with a significant amount of money needed for maintenance and network usage. However, because cellular operators would be able to use such a buildout to improve their own coverage, cost sharing could slash the investment needed from governments and local authorities. How the communications companies and government split those costs will likely prove crucial to how smart roads get by 2050.