Clarification

• Do we need to cover people taking public transportion such as bus or train? [No]
• Do we need to cover bikers or pedestrians? [No]
• Do we need to cover taxies?[Yes]
• Do we need to cover service vehicles such as truck, USPS, Fedex? [Yes]
• If 2 people sit in a car waited 2 min at a traffic light, how much wait time is that? [4 min]
• Are we asking about the daily wait time or yearly? [Yearly]

• Users 0-5 yrs old
  • US has 320M population, 80yr life expectancy.
  • ~20M in this segment
  • 40% 2 car trips/day as passenger
  • 60% 0 car trips/day
  • This gives about 16M trips/day
• Users 6-18 yrs old
  • ~48M in this segment
  • 20% 2 car trips/day as passenger
  • 80% walking, biking or school bus.
  • This gives about 20M trips/day
• Users 18-22 yrs old
  • ~8M in this segment
  • 10% 4 car trips/day
  • 20% 2 car trips/day
  • 70% 0 car trips/day
  • This gives about ~6.4M trips/day
• Users 23-65 yrs old
  • ~168M in this segment
  • 10% 10 car trips a day, such as taxi drivers,
  • 30% 4 car trips a day
  • 50% 2 car trips a day
  • 10% 0 car trips a day, walking or taking public transportation
  • This gives about ~500M trips/day
• Users 65+ yrs old
  • ~60M in this segment, mostly retired
  • 40% 2 car trips/day
  • 60% 0 car trips/day
  • This gives about 48M trips/day

• 80% of trips are through freeways
  • No traffic lights in freeways
  • 0~3 miles to get to freeway entrance
  • 0~3 miles to get to destination from freeway exits
  • 1 traffic light every 0.3 miles
  • 50% of chance you need to wait.
  • This gives about 1.5/0.3*50%*2=5 waits/trip
• 20% of trips are through local roads
  • 0~3 miles
  • 2.5 waits/trip
• overall it will be 4.5 waits/trip

The Google estimation question: How much time is spent by people while waiting in their vehicles at traffic signals in the US.

Clarifying question

1.       When you say people waiting in their vehicles do you mean the driver or the driver plus all the passengers in the car.  I’d prefer to focus on the driver and not worry about passengers.

2.       When you say their vehicles can we focus personal transportation not busses and other multi-person transport or business driving?  I’d prefer that limit.

3.       When we talk about vehicles technically bikes are included under the vehicle code – but I’d prefer to focus on motorized transport?

4.       When you say traffic signals can we limit that to traffic lights, not train stops, stop signs etc.

5.       Can I base the assumption on pre covid driving?  To be honest I haven’t driven much since march, and I don’t have much of a frame of reference for current driving trends.

6.       I’m going to base my assumption starting with US population this won’t account for tourists is that OK?

Framework

I’m going to start with the us population

I’m going to reduce that to household

I’m going to reduce households to cars.

I’m going to estimate the number of miles driven per car

I’m going to break the miles between freeway with low or no traffic lights and surface streets with traffic lights

I’m going to estimate the number of traffic lights that the average driver hits per mile.

I’m going to estimate the wait time

Then I’m going to sum everything up

Does that framework make sense?

Estimation

The US population is 330,000,000

Households reduce that by 2.5 = 132,000,000 – 2.5 people per household is a stat that I have heard many times before

Cars reduce that by 1.5  –  88,00,000 – Not every household has a car – this 1.5 is a stat that I have heard before.

The oil lube industry is entirely based on the average car going about 3000 mile per 3 month.  This number also jives my experience getting car insurance where they except you to drive about 12 miles per year.  I also believe that the leasing industry is based on about 10K per year and I think they set this a bit low so they can charge you on the back side or use it as a negotiation point for renewal.  Given those data points I think this is a good estimation.   I am confident that some people drive a lot more and some people drive a lot less – hopefully across this size population it will average out.

That means we drive about 1,056,000,000,000 miles.

Some of that driving is on the freeway and some of that driving is not.  My frame of reference is that about ½ the time is freeway and the other ½ is not.  I could try to breakdown drivers into different persona but I think the data set is large enough that it will average out.

528,000,000,000 freeway

528,000,000,000 not freeway

Not all, not freeway driving is the same there is driving in suburbs, city and rural.  I could attempt to break that down City drivers would travel less miles city’s have more public transportation users and hit more lights, rural drivers would be the opposite and suburban drivers would be in the middle.  I think if we start breaking this down we would be in total spitball territory,  so again unless you want me to I’m going to assume that it all washes out in the big data set.

When I goto the big box stores near me it’s about a 3 mile drive and I might hit 1-4 traffic lights.  Lets go with the average of 2.5 traffic lights per 3 miles.

A traffic light duration I believe is about 2 min – but sometimes you get through fast and some time you wait the whole time – lets split the difference and say 1 min is the average wait time.

This means based on my estimates the not freeway portion of the driving would be

(528,000,000,000/3)*2.5 = 440,000,000,000 minutes

The freeway portion generally doesn’t have lights except when it’s metered at rush hour.  (now you can’t do this by lights per mile – it’s lights by trips during metered.  We can spitball this (really spitballed)  the average freeway trip is 20 miles (based on the very small sample size of my experience) and I hit the meter about 20% of the time.  We will use the same light duration for the meter – they go faster but sometimes the back up so still 1 min.

(528,000,000,000/20)*.2 = 5,280,000,000 minutes

Total estimated minutes 440,000,000,000 +5,280,000,000 = 445,280,000,000 minutes or 12.6million hours.

Limitations

There were a couple of big assumptions made here that we should know about.

1.       Population to household should be solid

2.       Household to cars is a good faith guesstimate – in fact the number of cars registered in the US apparently outnumbers licensed drivers  https://en.wikipedia.org/wiki/Passenger_vehicles_in_the_United_States

3.       The miles driven is actually pretty solid – the actual number is 13.5K

4.       The split between freeway miles and not freeway miles is a guess based on the exceedingly small sample size of me

5.       The number of lights hit per 3 miles is based on my suburban life

6.       The wait time is solid – the actual cycle 2 min.

7.       In retrospect maybe working from licensed drivers might have been a better approach then by car.  But… that is estimation for another day.