Remote Operator

Baby, You Can Drive My Autonomous Car

Both GMGM and AlphabetGOOG are planning to launch autonomous taxi services this year.1 Because autonomous vehicles will not be able to drive without human help 100% of the time, a remote operator will have to assist. Before autonomous taxis commercialize, professionals probably will drive or direct vehicles remotely, especially during emergencies. According to ARK’s research, autonomous taxis could cost consumers 35 cents per mile, 5 cents of which will go to the remote operator network.2

While beaming video at low latency across cellular networks has been challenging,3 several startups have been working to find solutions. Phantom Auto, for example, bundles multiple cellular networks and sends prioritized snapshots of live video feeds to a remote operator who can direct, or tele-operate, vehicles with a set of brakes and a steering wheel.4 With multiple cellular networks, Phantom can deliver its tele-operator service at acceptable latency levels.5 Other startups working on remote operations include Alphabet-funded Scotty Labs, and Starsky Robotics, which is focused on the remote operation of autonomous trucks.6 At the Consumer Electronics Show this year, Nissan also announced a remote operations center, one of many likely as autonomous vehicles commercialize.

ARK estimates that the remote operator network will cost approximately 5 cents per mile, including the operator and the cellular network. At more than 15% of the overall cost, it will be crucial to a near-seamless taxi service in which the passenger is unaware of any challenges. Worst case, an autonomous car will be able to pull to the side of the road and receive instructions, a small price to pay to ensure a safe and inexpensive ride.

ARK believes remote operator networks could be a massive windfall for the telecom industry. If the cost of data were to remain at current levels, ARK estimates that remote taxi service operations could add $100 billion, or 40%, to telecom revenues in the US.7 Globally, telecom revenues could roughly double from today’s $2 trillion market.8 More likely, the cost of data transmission will drop precipitously over time, but still will benefit global telecom revenues by $460 billion if the global average cost of data were to drop by 75%.9

The introduction of 5G cellular could speed video feeds from autonomous cars over cellular networks by orders of magnitude. As a result, remote operations centers should receive more accurate and detailed video feeds and real time maps resulting in more immediate updates. Although 5G rollouts will not determine the timing of commercial autonomous taxi launches, they could improve service and reliability levels and push prices down to levels much lower than we have anticipated.


Banner Image Uploaded by Andrey Suslov

  2.  ARK Investment Management LLC, 2018. ARK is estimating that an autonomous taxi network would include thousands of cars, with each car requiring help 1% of the time.
  5. Note that Phantom Auto does not disclose what level of latency they are able to achieve.
  7. ARK Investment Management LLC, 2018;
  8.  ARK Investment Management LLC, 2018;

    Notes: Two notes about the discrepancy between the effect of autonomous taxis on US and global telecom revenues. 1) The percentage of the population that is currently in the point to point mobility market in the US is much higher than the global average, but ARK estimates that when autonomous taxis make point to point mobility much cheaper than today’s prices, this will encourage many more people into the market, driving up data needs for autonomous taxi services. This demand increase will be much more profound in the developing world. 2) Today the US population consumes much more data than the global average, which allows for cheaper domestic data pricing. However, global consumers still have a similar need for getting place to place, so we don’t expect the same consumption patterns to be mimicked in the autonomous taxi market.

  9.  ARK Investment Management LLC, 2018

ARK's statements are not an endorsement of any company or a recommendation to buy, sell or hold any security. For a list of all purchases and sales made by ARK for client accounts during the past year that could be considered by the SEC as recommendations, click here. It should not be assumed that recommendations made in the future will be profitable or will equal the performance of the securities in this list. For full disclosures, click here.