Posted by MIRA Funds
26 Feb 2020
1.35 million people die on the world’s roads each year1 with approximately 95% of accidents caused by human error2. With congestion also costing $US300 billion each year to the US economy alone, automotive manufacturers, road users and infrastructure managers are looking to AVs as a means to improve road safety and efficiency.
All AVs are not created equal.
According to a classification system devised by The Society of Automobile Engineers, AVs fall on a spectrum of zero to five. ‘Level Zero’ vehicles are completely manual, requiring drivers to perform all tasks whilst driving. At the other end of the autonomous spectrum, ‘Level Five’ AVs are entirely automated – requiring no human intervention whatsoever whilst on the road.
Although much of the attention paid to AVs is focused on vehicles at the upper end of the autonomous spectrum, ‘Level Three’ technologies from parking assistance to cruise control have been in the mainstream for a number of years. Despite limited pilot programmes involving ‘Level Four’ robotaxis having been launched in the US and Singapore, ‘Level Four’ and ‘Level Five’ AVs face a number of hurdles before they can be deployed more widely.
One of the most significant challenges will involve satisfying the high safety standards expected by regulators and the community. AVs collect and process significant amounts of data in order to understand and interact with the vehicles and infrastructure around them. This technology must be developed to handle so-called ‘edge cases’ – situations where a vehicle encounters something it has not experienced before and does not know how to react to.
Although reaching this stage of development is not impossible, car makers are coming to appreciate the challenges involved in making AVs capable of handling all situations they may face when on the road.
All major automotive manufacturers have recognised the importance of AVs to the future of transport. The question for carmakers is, “When will a market exist for my AV?”
The advent of AVs could create a market worth $US7 trillion by 20503. However, the remaining technological challenges and uncertainty about if or when robotaxis will become a genuine reality has made forecasting AV take-up difficult, with current AV adoption predictions varying wildly. For example, as noted in the table below, UBS’s conservative case for AV sales by 2030 is just 2.5 million, compared to its bull case of 40 million.
In other areas automation technology is, however, already a commercial reality. Automation in the agricultural and port handling sectors is, for example, already in use, reducing costs and adding to efficiency in these sectors.
AV sales in 2030 as a percentage of total automobile sales.
|Company||Units (million)||% of total sales|
|McKinsey – Base case||16.0||15.2%|
|UBS – Base case||12.6||12.0%|
|UBS – Bull case||40.0||38.0%|
|UBS – Conservative case||2.5||2.4%|
Source: Bloomberg New Energy Finance (BNEF), Long-term AV Outlooks (May 2018)
Given their convenience, privacy and comfort, robotaxis have the potential to impact demand for public transport such as buses and metro services. In reality, however, any significant shift to AVs in peak commuter hours could result in unacceptable levels of additional congestion. In such cases, local municipalities may impose surcharges – providing respite for buses, trams and trains.
Automation should mean that more cars will travel along roads and across bridges, given the opportunity for; narrower lanes, the ability for cars to travel closer together, and fewer accidents. When combined, these factors could significantly increase the earning potential of these assets.
Overall demand for parking could weaken if robotaxis ultimately prove viable in urban areas. However, car park operators may also receive significant benefits from automation technology. Self-parking systems will mean utilisation and revenue per square metre could increase. Car park operators will also be able to offer additional services to robofleet companies, such as cleaning and charging. These potential opportunities for margin and revenue improvement should offer comfort in the face of some demand headwinds.
AV technology is already in use in the form of automated baggage handling systems at airports and automated guided vehicles (AGVs) in seaports. These environments – which are far less complicated than conditions faced by AVs on the open road – are well suited to currently available AV technology. More widespread adoption could see significant cost and efficiency gains for these assets as they integrate autonomous technology into their operations.
Given the number of sensors and recording devices they rely on, AVs are expected to capture vast amounts of data. Some reports suggest this will be as much as 30 Terabytes per vehicle per day – equivalent to 3,000 times the amount of data produced by all of Twitter’s 270 million users in a day combined4. Industry forecasts predict that due to the proliferation of connected devices and cloud computing in general, the demand for data centres is expected to increase significantly over the next decade5.
While an AV does not necessarily need cellular access to function, the success of many of the Vehicle-2-Network (V2N) applications are dependent on access to a reliable cellular network. As such, there could be opportunities for mobile network operators to offer enhanced services to mobility providers.
AVs are an exciting technology in many respects. A strong understanding of the technology, its likely adoption timeline, and its implications, could unlock a number of opportunities for infrastructure investors.
While there is a lot of optimism about the timeline for AV adoption, we believe full ‘Level 5’ (or even wide-ranging ‘Level 4’) adoption is still some way off. With residual edge cases proving particularly difficult to solve, as well as legal and regulatory concerns still to be worked through, we do not expect to see broad-based AV adoption prior to 2030. It could be closer to 2040 before human drivers are in the minority.
1 The World Health Organisation February 2020, https://www.who.int/en/news-room/fact-sheets/detail/road-traffic-injuries
2 Royal Society for the Prevention of Accidents (RoSPA) 2017, https://www.rospa.com/rospaweb/docs/advice-services/road-safety/road-crashes-overview.pdf