A Financial Analysis of Different Scenarios for Using Autonomous Vehicles to Deliver Packages


As the technology to automate road vehicles and the policies regulating them continue to advance this paper explores potential ways that it could be integrated into the delivery industry. The paper makes two contributions. First, it proposes different ways that autonomous delivery vehicles – the vehicles responsible for carrying packages from a sorting center to the customers – could be designed. Second, profit and loss (P&L) models are built using publicly available data about delivery company operations. These models are used to calculate the minimum performance levels that these vehicles would need to deliver, under different cost and pricing scenarios, to keep profitability for delivery companies at existing levels. The two scenarios proposed are termed self-driving trucks (SDT) and locker-on-wheels (LoW). SDT are used to enhance the productivity of couriers by freeing them to locate and collect packages while the vehicle is in motion. The saved time allows couriers to deliver more packages to customers. LoW lack a driver which reduces cost. These vehicles travel to customers and require self-service collection. Kamin & Morton – A Financial Analysis of Different Scenarios for Using Autonomous Vehicles to Deliver Packages

AUSVI – Valuing the Convenience of Autonomous Vehicles

SMART – AVS -Session 1 Abstract


Fully autonomous self-driving cars (SDC) must overcome numerous policy hurdles before they can be made commercially available. To ensure these issues receive proper attention it is important that the societal benefits of this technology are well documented. SSDC (dense urban environment) PSDC (all other environments) Book + Wait vs. Walking -109 23 Parking Search 115 90 Walking from Vehicle 62 54 Servicing, Fuelling, Cleaning 20 20 Total 88 187 Note: The value of time is also changing. So, even though a person spends more time waiting on an SSDC, this time can be spent in comfort, thus mitigating the social cost. Economic Development Research Group. Economic Impact of the Massachusetts Turnpike Authority & Related Projects – Volume 1 (2006) Eno Center for Transportation. Preparing a Nation for Autonomous Vehicles (2013) In this work, we expand on the methodology for valuing time used by transportation planners. We do this in two ways. 1) We look beyond on-road travel time and account for the value of the time saved parking and walking. 2) We account for the value of being able to engage in work or leisure while traveling. The resultant value of travel time savings (VTTS) is more than five times larger the often cited safety benefits of this technology. The combined impact presents an even more compelling case for government policies and investment.

Toward a Systematic Approach to the Design and Evaluation of Automated Mobility-on-Demand Systems: A Case Study in Singapore

Toward a Systematic Approach to the Design and Evaluation of Automated Mobility-on-Demand Systems: A Case Study in Singapore

Citable URI: http://hdl.handle.net/1721.1/82904

Department: Massachusetts Institute of Technology. Dept. of Aeronautics and Astronautics; Singapore-MIT Alliance for Research and Technology Center.
Publisher: Springer
Date Issued: 2014-04

The objective of this work is to provide analytical guidelines and financial justification for the design of shared-vehicle mobility-on-demand systems. Specifically, we consider the fundamental issue of determining the appropriate number of vehicles to field in the fleet, and estimate the financial benefits of several models of car sharing. As a case study, we consider replacing all modes of personal transportation in a city such as Singapore with a fleet of shared automated vehicles, able to drive themselves, e.g., to move to a customer’s location. Using actual transportation data, our analysis suggests a shared-vehicle mobility solution can meet the personal mobility needs of the entire population with a fleet whose size is approximately 1/3 of the total number of passenger vehicles currently in operation.
Citation: Spieser, Kevin, Kyle Ballantyne, Treleaven, Rick Zhang, Emilio Frazzoli, Daniel Morton, Marco Pavone. “Toward a Systematic Approach to the Design and Evaluation of Automated Mobility-on-Demand Systems A Case Study in Singapore.” Forthcoming in Road Vehicle Automation, Springer Lecture Notes in Mobility series
Version: Author’s final manuscript
Terms of Use: Creative Commons Attribution-Noncommercial-Share Alike 3.0
Journal: Forthcoming in Road Vehicle Automation


Why Singapore Should Go Driverless (The Straits Times March 2, 2013)

Why Singapore Should Go Driverless

Why Singapore Should Go Driverless

The Straits Times | March 2, 2013 by Daniel James Morton

As a long time admirer of Singapore, my bold yet feasible proposal is that Singapore deploy the first city-wide driverless car network. This will spawn a trillion dollar Self-Driving Economy that dwarfs the current $2 billion vehicle tax revenue. Further benefits include improved mobility, greater safety, more energy independence, international prestige, new services and enhanced social unity. The country’s strong government institutions and advanced economic development provide the competitive advantage to initiate this transformation. Singapore can lead this change worldwide at no cost and derive a financial windfall larger than any existing industry in the country.

Click to download:   Straits Times March 2, 2013 : Driverless car could rev up Singapore Economy (pdf)

Google’s driverless car isn’t about revenue…well not advertising revenue

What is the value of the self-driving car for Google ?

How much is this car worth to Google ?

There are rumors that the “do no evil” company is working on the driverless car to generate advertisement revenue from drivers. Freed of the constraints of the wheel and navigation, drivers along with their passengers can instead immerse themselves in an online experience and thereby generate significant advertising revenue for Google, the company’s dominant source of income.

Google can increase its value $16 billion by delivering advertisements to the Self Driving Economy in the US, its largest market. This represents only slightly more than 6% of its $230 billion market cap. This is not the driver behind Google’s ambitions. Let’s dispel the myth further below. Though as an aside, if Google did generate this advertising revenue from drivers, the billboard advertising market of 7BN per year[1] would disappear.  Who will be looking at billboards when everyone’s eyes shift from looking forward to looking down? Google’s financial opportunity may lie in its ability to provide the navigation services to the Self Driving Economy (a future post).

Under the hood: How long do we spend in cars?

On average, people in the US over 18 are spending 962[2] hours per year in the car. There are 237[3] Million people over 18 in the US representing 229 billion hours per year spent in the car.  People under 18 also spend time in cars and use internet devices. There are 75 million[4] people under 18, we then subtract out the age range 5 and under and get 54 million. Assuming they spend the same amount of time in cars per year as the over 18, they spend 52 billion hours. Combine these 2 groups and there is a total of 280 billion hours spent in the car per year in the US. I hope you like your seat.

Not all these hours are spent driving since this estimate ncludes passengers. We could estimate average occupancy to calculate driving hours. A reasonable proxy is available for private transport commuters. It is a lonely experience, 76%[5] of commuters are driving alone. Yet commuting only represents about 20%[6] of all miles driven. However, I contend that freeing up the driver also frees up the passengers to consume more online media. Passengers often socialize and help the driver. Furthermore, cars are not currently designed to have a broadband network inside since the driver is driving. A portion of passengers could already be using online devices while in the car yet to err on the high side we will assume that all hours spent in the car (280 billion hours) are now available to consume online media.

How many of these hours might be spent using Google?

Google users spend about 1:54[7] hours per month on Google. They spend a lot more time on Youtube (a Google property), about 6:52[8] hours per month. At just under 9 hours per month on Google + Youtube, this represents about 1.2% of all available hours (yes including sleeping). Almost all activities performed by people in their daily life, categorized by the US household survey, can be performed also while in the self driving car i.e. working, sleeping, socializing. A few items cannot be performed such as home chores. Therefore if we back out the time spent on household chores (and caring for others) which could not be performed in the car, we get a slightly higher figure of 1.34%[9].

I assume that a Google user will allocate this same % of their time to their newly freed up time in the car. A few items will effect this average both up and down such as car sickness (you wouldn’t want to read but you might want to watch more videos), internet access (you can’t always surf/stream so more offline usage), preference of video watching over other leisure activities, faster commuting time will be offset by more time to use google outside the car so it is a wash. I will assume these factors don’t materially change the average usage so I maintain the 1.34%.

Previously, we estimated that people spent 280 billion hours of time in a car per year in the US. This includes people Google users and non Google users. Google has 175 million unique users in the US.[10] This represents 60% of all people over 5 years old (5 years old watch “Charlie bit my finger” and generate ad revenue). Assuming 60% market share and that Google users don’t spend more or less time than the average person in a car, then Google users spend 168 billion hours per year in the car. If they allocate 1.34% to using Google and Youtube this creates an additional 2.3 Billion hours of ad opportunity for Google.

How much value does Google generate per hour in the US per user?

Forbes estimated that Google generated $0.35[11] per hour for a global user. My estimate for a US user is several times this, around $1.15. This is based on 1st 9 months of 2012 revenues of $15.8bn[12] which annualized is $21bn. The amount of hours per year is approximately 18.4bn broken down into 14.4bn for Youtube (4bn per month remove the 70% outside the US and annualize)[13] plus the 4bn for Google (175million users at 1.9 hours per month annualized). The Google+ usage figures of 12 minutes per day for active users refers to the social networks users usage across all Google properties so we can’t add it or it would be double counting.[14]

The ads in the car would allow for location-based advertisement which could increase the value proposition to suppliers. However, this shift has already started with mobile and tablet location based ads which are priced less (16% in one estimate)[15] than for desktop computers. Therefore, $1.15 per hour is on the high side as the shift continues from desktop to mobile and in-car usage. On this $1.15 per hour in revenue, I assume a 25.7%[16] net income margin.[17] At $0.29 net income per hour per user in a car we take the P/E of 24.19[18] and estimate that each Google US user hour in the car increases Google’s value by $7.12.

With 2.3bn additional hours per year this results in an increase of $16bn in market cap which is 6.7% of its current market cap of $240bn (fluctuating daily of course).

This considers only the US but other countries spend far less time in their cars, take the UK[19] for instance (the only non-US geography split out in Google’s financial statements) where they spend just over half (54%) of the time someone in the US spends. At ½ the time, and 1/5 the addressable market[20] and assuming same ability to monetize, this would represent 10% of the US additional value.

Conclusion: The extra viewing time on Google properties will generate more value to the company but there could be bigger opportunities. How much value could Google generate providing navigation services to the Self Driving Economy?

[1] http://www.ibisworld.com/industry/default.aspx?indid=1437

[2] http://www.arbitron.com/downloads/InCarStudy2009.pdf

[3] http://quickfacts.census.gov/qfd/states/00000.html

[4] http://www.census.gov/population/age/data/2011comp.html

[6] http://www.census.gov/prod/2011pubs/acs-15.pdf, n.b. this is miles driven not time spent driving but we will assume that there is not a material difference in average spend when commuting versus non commuting.

[7] http://blog.nielsen.com/nielsenwire/online_mobile/september-2012-top-us-web-brands/

[8] http://www.youtube.com/t/press_statistics – this reconciles closely to comscore’s estimate of 6 hours and 28 minutes for Google properties – http://www.comscore.com/Insights/Press_Releases/2013/1/comScore_Releases_December_2012_U.S._Online_Video_Rankings

[9] http://www.bls.gov/news.release/atus.htm  n.b. transportation time is included in each activity as part of the time spent and not broken out separately.

[10] http://blog.nielsen.com/nielsenwire/online_mobile/september-2012-top-us-web-brands/, n.b. this includes under 18 year olds who also use google.

[11] http://www.forbes.com/sites/georgeanders/2012/07/26/facebook-ratio/

[12] Excludes Motorola revenue

[14] http://marketingland.com/google-users-spend-12-minutes-per-day-in-the-stream-15423

[15] http://www.nytimes.com/2013/01/23/technology/google-profit-exceeds-expectations.html?_r=0

[16] http://www.sec.gov/Archives/edgar/data/1288776/000119312512440217/d400524d10q.htm

[17] Again, exclude Motorola

[18] http://ycharts.com/companies/GOOG/pe_ratio

[19] http://www.sheilaswheels.com/media/news/BRITS_SPEND_MORE_TIME_DRIVING_THAN_SOCIALISING.html

[20] http://www.ons.gov.uk/ons/taxonomy/index.html?nscl=Population

Why is Car 2.0 not yet happening? It is all about incentives.

The universal adoption of Car 2.0 is now a financial equation. The financial gains to society as a whole will be tremendous. However, as with any technological disruption, there are existing players that will adapt and benefit, while others will cease to exist. Car 2.0 is so transformative that this may be one of the biggest Blue Ocean opportunities. The emergence of Google as an actor in the long established car industry and leading the lobbying efforts in the US is indicative of the change to take place. Understanding the motives and incentives of all the individual actors is a complex affair.

Winners and Losers

The seemingly obvious innovators to lead this transformation are the car manufacturers. Car 1.0 is lead by mechanical engineers whose companies are designed around sourcing and manipulating heavy parts to build mechanical cars weighing 2 tons. Car 2.0 is lead by software programmers and designers. The transition to Car 2.0 puts the existing car companies existence at risk since part of their assets become liabilities. Furthermore, they face the chicken and egg problem. They cannot commercialize an optimized Car 2.0 (i.e. lightweight) into a transportation eco-system based on metal armor. This codependency hampers any individual car company from breaking out. A small controlled environment such as Singapore is an ideal place to launch.

The Car as a product will change completely making part of the expertise and assets of existing car manufacturers obsolete. The Car 2.0 services industry will also emerge and will become larger than the Car 2.0 product industry. At the onset of the computer revolution, hardware was king and then the value shifted to software companies and then the services industry. A similar shift will occur in the car industry.

The shift to Car 2.0 will also lead to a change in product liability risk. All car companies have initiatives to reduce accidents through increased driver assisted functions (self-parking, lane changing, automatic breaking etc). This shift in control (and hence responsibility) from the driver to the car comes also with an increase risk in product liability to the car manufacturer. Car manufacturers in both the US and Europe face this challenge. The difference is that the potential liability in the US is several orders of magnitude larger than in Europe. As a consequence, car companies are less incentivized to innovate since they can only sell the cars to a portion of the market. The unquantifiable financial liability that the car manufacturers might expose themselves to is a show stopper. Both regulators and insurance companies need to come to the rescue but will they?


Given that succeeding in the Car 2.0 era requires a different set of skills and assets than Car 1.0, the car manufacturers are not totally aligned today with developing Car 2.0. In developed countries, car manufacturers along with their large workforce and political ties have a dominant status in the economy. They have not as of yet used their political might to advance the Self Driving Economy. Many well informed automobile experts refute that Google has changed the playing field and cite that automobile companies have embarked on automating cars for decades. The issue is not if the automobile companies have the technical knowledge and capability, it is about incentives. Do the automobile companies have the incentivise to innovate in this direction? At present most of their efforts are on incremental solutions (automated parking, lane change…) so as to be accepted by consumers and economically viable for them in the short term. As a consequence, the regulators are not lobbied to advance the transition. Nevada was the first US state to pass a law to “start” to regulate driverless laws. This was pushed through by Google. Why didn’t this occur in Michigan in the homeland of the main car manufacturers? Regulators will not lead this movement and given the powerful interest groups that will seek to slow down any progress (think Teamsters in the US) that they will be a lagging indicator.

Insurance companies…

The insurance companies could provide a private solution where the public regulators cannot. Their role is to price the risk. This allows existing car companies and new entrants the financial basis to make a risk/reward assessment. However, the interests of the insurance companies are also not aligned with the Car 2.0 industry. If the number of car accidents is reduced significantly then risk decreases, hence car insurance premiums must be reduced and insurance companies profits reduced. During the transitional period, some insurance companies will overcharge for “new coverage” since they will be early entrants and generate windfall gains. However, this opportunity will be eventually arbitraged out as the premiums are reduced to reflect the real reduced risk. Car insurance companies will see their business challenged.

10 Personal Benefits from the Driverless Car

Greater safety for your family

Top personal benefit : greater safety

The deployment of a driverless car infrastructure and the redesign of the car will lead to a wide range of personal benefits:

  1. Greater Safety : Car 2.0 can drive more safely than people (Car 1.0) resulting in less deaths and accidents and less worry for non-drivers i.e. pedestrians and cyclists
  2. Proliferation of access to shared and private transport options on a pay-as-you-go model (imagine driverless taxis and driverless buses) thus solving the public transportation challenge
  3. Increase in personal productivity and leisure time since people are no longer driving but chauffeured
  4. Shorter and more predictable travel times since greater driving precision reduces congestion exponentially and passengers no longer need to find parking spaces since Car 2.0 park themselves
  5. Greater choice in deciding where to live thus reducing housing costs and commuting stress
  6. Dramatic increase in flexibility and new services: Car 2.0 can perform home or office deliveries or be summoned to pick you up
  7. Explosion in job creation across a range of expertise with the birth of the Car 2.0 industry
  8. More enjoyable, safer and healthier personal modes of transport such as bicycling, jogging, walking, and rollerblading
  9. Improved air quality since car emissions will be drastically lowered
  10. Cheaper transportation costs with greater personalized quality.