Autonomous vehicles: Billions needed

Fully autonomous vehicles (AVs)—those at level 4 or 5 on the SAE scale—require extremely complex software and hardware integration, yet there is no shortage of companies working on the problem. More than 50 companies worldwide are testing AVs, while many others are developing the underlying software and hardware. Several dominant technology companies with virtually unlimited balance sheets, are also racing to establish a first-mover advantage. The pace of investments in AVs in the last several years has accelerated, with $61 billion in total spending as of June 2018, and $4 billion in the last 12 months alone. Some of the notable investments include GM's Cruise Automation (approximately $4 billion) and Intel's Mobileye (approximately $15 billion).

We analyzed disengagement data (the number of miles an AV can drive before human intervention is required) from the California Department of Vehicles, as shown in figure 1. That data shows that Alphabet's Waymo and GM's Cruise Automation are leading the pack, with many others still in the early stages of development. Yet with complex software, mapping, and hardware integration required, we believe billions of dollars in incremental investments will be required. Clearly, not all of the 50 companies developing autonomous vehicles will emerge as winners, and most of the investments made by others will not pay off, with perhaps only a few companies expected to win in this market.

Our 2018 Autonomous Vehicle Survey of US consumers shows that 19% are likely to consider purchasing an autonomous vehicle, with a much higher number of consumers (57%) not interested, as shown in figure 2. In addition, the same survey of consumers estimates that the willingness-to-pay (WTP) for level 4/5 AVs is $2,300 per vehicle. (Among early adopters—defined as the 19% of consumers likely to purchase an AV—the willingness-to-pay increases to $3,300.) Our estimate of the incremental cost of a level 4/5 autonomous vehicle system is 10 times higher ($22,900 per vehicle as of 2017); even in 2030, our forecast of the incremental cost is $2,600 for a level 4/5 system. Therefore, commercial applications (e.g. robotaxis) will lead deployment and the industry will need to drive the cost down by more than 80% in order to meet the price expectations of the majority of consumers.

How will this impact global automotive OEMs and suppliers?

• The disengagement data shows only a couple of companies (Waymo and GM's Cruise Automation) with comparatively higher levels of technological maturity. As further investments are made to develop capable AVs, most of the other 50 companies working on software and hardware will not emerge as winners.
• As companies race to gain a first-mover advantage and sharply ramp up the investments, industry leaders face a key question: how can they ensure that the billions they're investing in AV development are not wasted?