Dynamic Bridge Tolls

On my way into work today, I was thinking about bridge tolls. There has been talk of raising the tolls on various Bay Area bridges to $4 or $5 or even $6! Yikes!

Now that we have FastTrack, it seems that there may be more creative solutions we can come up with. For instance, what if the toll is variable? Those crossing at 5am get to cross for $0.50. At 6am, its $1.5, at 7am it jumps to $3, at 8am its $6, at 9am it drops to $3, etc etc?

I found a similar proposal for the New York area from the Tri-State Transportation Campaign which had some good data in it here. They divided their plan into “off peak” and “on peak”, at $4 and $6 respectively.

Using excel, a little math, and a bit of arrogance, I am coming up with this model. My goals are to increase revenue, but also increase vehicles-per-hour over the bridge.

1) If you don’t use fasttrack its $7.50. This is because you are costing us a lot of money and slowing everyone down. Get fasttrack.

2) Peak hours (7am to 9am): $7

3) Off-Peak hours: $2

Now this is a no-comprimise plan with teeth! A commuter can save $25 per week by shifting to off-peak driving hours.

Using the same percentages for vehicles per hour as the TSTC used in New York, my model would produce a 5.7% increase in revenues over the current, static $3 plan. At the same time, it puts such a heavy incentive to get drivers onto FastTrack and to not use peak hours, I think we’d see a significant decrease in overall bridge wait times.

But what do I know….

Really Extreme

But what if we wanted to get really radical, and have a completely variable price that adjusts in real time?

Google is an interesting example of a company that adjusts their prices in real time. They deal with a huge volume of “ads”, and they are constantly figuring out which ad is best optimized for their business. This is done in real time.

What if all California bridges were linked electronically. As tolls are collected (either digitally via FastTrack or via toll-takers), the central system adds up the money. At the same time, the central system knows how many vehicles per hour are flowing through any bridge.

Each year, the governor types in a number for how much revenue per day he wants to get from the Bridge system. Then, the bridge system uses its historical models, current traffic flow rates, and the target revenues to calculate how much it should be charging drivers to cross the bridge at that instant.

Imagine the case where a bridge is extremely congested. The bridge system would automatically start raising the toll – in REAL TIME. This would have the effect of also lowering the tolls on the other non-trafficked bridges also in REAL TIME. We could set limits on the toll so that it never exceeded say, $15. But when it gets congested, commuters need to find another way to get home.

The advantage of this system would be that bridge tolls would be almost free when crossing if there was no traffic. But, when congestion occurs, the price would get higher and higher. Its actually a very fair way to deal with pricing.

Anyway, this idea is wild and crazy. The state of California isn’t very capable for creating real-time, dynamic bridge crossing toll algorithms. And we can get “close enough” with something simpler and less politically damaging anyway…

Leave a Reply

Your email address will not be published. Required fields are marked *