Improving Truck Assignment

10 posts / 0 new
Last post
Alex Bettinardi
Improving Truck Assignment

I have a transportation assignment question for the wild world as 2020 comes to a close.

I'm wondering what ideas / treatments have been applied to try and keep trucks on major freight corridors as opposed to picking an optimal path by simply time, distance, and cost.

As background, this is for Oregon's Statewide model, where we assign freight and passenger traffic separately.  Over the years we have tested various weights of distance versus time, but neither really solves the issue.  In either case (distance priority or time priority) there can be odd network configurations where a shorter time or distance path is through a neighborhood versus on the freeway.  So if we shift the priority some sections look better as far as attracting the right amount of freight traffic and some sections look worse.  Shift it again, and different areas look better and different areas look worse.  To be clear, we block links where trucks are not allowed to travel (and this helps), but there are many links where trucks can travel to reach their shipping destination but they are not primary routes, they are just side routes to finalize the last mile of delivery.  But assignment doesn't know that and uses them at the first instance of those routes becoming optimal.

We recently had a peer state give us an interesting idea - pre-load the assignment with trucks first, so that trucks make a longer trip length decision without knowledge or the easy ability to shift based on local congestion changes.  This idea was very interesting, but after looking at it more, it seems like that would be problematic for analysis in Oregon.  Specifically as Oregon continues to move forward with tolling analysis.  If trucks see a congestion free network first, they would be the first to jump ship at a policy made bottleneck like a toll.  So where some might expect trucks to be the last vehicles to try and avoid the toll; if they are assigned first, they would likely be the first (most likely) vehicles to re-route around the toll.

My current plan/thought is to tag links with national freight highway designation.  Any link tagged as a freight route would see the normal link impedance in assignment (1x).  All links not tagged as a freight route would have impedance multiplied up by something like 2x (would need to play with the multiplier).  This would allow trucks to dodge tolls on other major freight routes, but it would put a high penalty on dodging a toll through a small commercial street.  It would also push the trucks to get on freight corridors as fast as possible and stay there for as long as possible - which I believe would be pretty close to reality.  

A long question for the end of a long year.  Please share your experience and thoughts with me.  Many thanks

Alex Bettinardi

John Gibb

Alex, first off, modeled trucks routing through neighborhoods is a potential diagnostic of network and/or assignment problems, along with whether car volumes are also overloading those neighborhoods.  Secondly, most pre-load assignment schemes date back to when modelers had to assign with HOV lanes before the software packages caught on to multi-class equilibrium.  All sorts of unexpected problems arise from priority-order assignment schemes.

As you seem aware, most packages now let you set up multiple user-classes distinguished not only by which users get to use which links, but also give different base (free-flow) speeds or impedances for each class.  (That's especially useful where there's tolling.)  I've worked with one model that gives trucks *lower* base speeds on freeways, reflecting their lower speed limits.  Another reasonable truck distinction is that heavy trucks travel slower than car traffic in local and urban settings due to frequent stops and turns, and slow acceleration, even if there's no congestion.  This seems a more realistic and less arbitrary basis than a blanket penalty, although it's hard to rule out a deterrence effect for going off the designated freight roads.  Good luck and happy new year.

Pedro Camargo

Alex,
It looks like you have already explored the problem a bit, but I
will add my 2c.

First of all, pre-loading truck trips sounds like a terrible idea and
reminds me of how things used to be done a couple of decades ago, and your
concern with policy testing is right on the money.

I would submit that going towards a more sophisticated generalized cost
function for trucks is definitely the best path, but I would suggest taking
a somewhat more sophisticated approach (assuming you have Truck GPS data),
and proceed to estimate a route choice model exploring all the variables
you may have (distance, free-flow time, congested travel time, road
hierarchy, legal restrictions, etc) in order to build or generalized cost
function.

Depending on your network and the OD pairs in question, treating for path
overlap might be needed and you would most likely have to adjust the scale
of the final generalized cost function when you bring it back to your
multi-class assignment, and while the latter is easy to deal with, the
former might be a little more laborious to implement.

You can also calibrate a cost function by trial-and-error, but that is both
less robust and a lot less fun. If you decide to go that route, I would
recommend looking into the truck route choice literature, as you might be
able to borrow some parameters from which you can start your exploration.

Again, just my 2c.

Cheers,
Pedro

On Sun, Jan 3, 2021 at 1:04 AM Alex Bettinardi <
alexander.o.bettinardi@odot.state.or.us> wrote:

> I have a transportation assignment question for the wild world as 2020
> comes to a close.
>
> I'm wondering what ideas / treatments have been applied to try and keep
> trucks on major freight corridors as opposed to picking an optimal path by
> simply time, distance, and cost.
>
> As background, this is for Oregon's Statewide model, where we assign
> freight and passenger traffic separately. Over the years we have tested
> various weights of distance versus time, but neither really solves the
> issue. In either case (distance priority or time priority) there can be
> odd network configurations where a shorter time or distance path is through
> a neighborhood versus on the freeway. So if we shift the priority some
> sections look better as far as attracting the right amount of freight
> traffic and some sections look worse. Shift it again, and different areas
> look better and different areas look worse. To be clear, we block links
> where trucks are not allowed to travel (and this helps), but there are many
> links where trucks can travel to reach their shipping destination but they
> are not primary routes, they are just side routes to finalize the last mile
> of delivery. But assignment doesn't know that and uses them at the first
> instance of those routes becoming optimal.
>
> We recently had a peer state give us an interesting idea - pre-load the
> assignment with trucks first, so that trucks make a longer trip length
> decision without knowledge or the easy ability to shift based on local
> congestion changes. This idea was very interesting, but after looking at
> it more, it seems like that would be problematic for analysis in Oregon.
> Specifically as Oregon continues to move forward with tolling analysis. If
> trucks see a congestion free network first, they would be the first to jump
> ship at a policy made bottleneck like a toll. So where some might expect
> trucks to be the last vehicles to try and avoid the toll; if they are
> assigned first, they would likely be the first (most likely) vehicles to
> re-route around the toll.
>
> My current plan/thought is to tag links with national freight highway
> designation. Any link tagged as a freight route would see the normal link
> impedance in assignment (1x). All links not tagged as a freight route
> would have impedance multiplied up by something like 2x (would need to play
> with the multiplier). This would allow trucks to dodge tolls on other
> major freight routes, but it would put a high penalty on dodging a toll
> through a small commercial street. It would also push the trucks to get on
> freight corridors as fast as possible and stay there for as long as
> possible - which I believe would be pretty close to reality.
>
> A long question for the end of a long year. Please share your experience
> and thoughts with me. Many thanks
>
> Alex Bettinardi
> --
> Full post: https://tmip.org/content/improving-truck-assignment
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13526
>

scott smith

Hi Alex,

In our Freight and Fuel Transportation Optimization Tool (FTOT), which routes over the national FAF network, we use an impedance penalty for the higher functional classes (e.g., 1.0 for Interstate, 1.3 for a local road). Since it is a national network, it does not include all links.

- Scott Smith

From: alexander.o.bettinardi=odot.state.or.us@mg.tmip.org On Behalf Of Alex Bettinardi
Sent: Saturday, January 2, 2021 10:02 AM
To: TMIP
Subject: [TMIP] Improving Truck Assignment

CAUTION: This email originated from outside of the Department of Transportation (DOT). Do not click on links or open attachments unless you recognize the sender and know the content is safe.

I have a transportation assignment question for the wild world as 2020 comes to a close.

I'm wondering what ideas / treatments have been applied to try and keep trucks on major freight corridors as opposed to picking an optimal path by simply time, distance, and cost.

As background, this is for Oregon's Statewide model, where we assign freight and passenger traffic separately. Over the years we have tested various weights of distance versus time, but neither really solves the issue. In either case (distance priority or time priority) there can be odd network configurations where a shorter time or distance path is through a neighborhood versus on the freeway. So if we shift the priority some sections look better as far as attracting the right amount of freight traffic and some sections look worse. Shift it again, and different areas look better and different areas look worse. To be clear, we block links where trucks are not allowed to travel (and this helps), but there are many links where trucks can travel to reach their shipping destination but they are not primary routes, they are just side routes to finalize the last mile of delivery. But assignment doesn't know that and uses them at the first instance of those routes becoming optimal.

We recently had a peer state give us an interesting idea - pre-load the assignment with trucks first, so that trucks make a longer trip length decision without knowledge or the easy ability to shift based on local congestion changes. This idea was very interesting, but after looking at it more, it seems like that would be problematic for analysis in Oregon. Specifically as Oregon continues to move forward with tolling analysis. If trucks see a congestion free network first, they would be the first to jump ship at a policy made bottleneck like a toll. So where some might expect trucks to be the last vehicles to try and avoid the toll; if they are assigned first, they would likely be the first (most likely) vehicles to re-route around the toll.

My current plan/thought is to tag links with national freight highway designation. Any link tagged as a freight route would see the normal link impedance in assignment (1x). All links not tagged as a freight route would have impedance multiplied up by something like 2x (would need to play with the multiplier). This would allow trucks to dodge tolls on other major freight routes, but it would put a high penalty on dodging a toll through a small commercial street. It would also push the trucks to get on freight corridors as fast as possible and stay there for as long as possible - which I believe would be pretty close to reality.

A long question for the end of a long year. Please share your experience and thoughts with me. Many thanks

Alex Bettinardi
--
Full post: https://tmip.org/content/improving-truck-assignment
Manage my subscriptions: https://tmip.org/mailinglist
Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13526

hudifebys

We have been facing the same issue with our regional and state(province)-wide models. In the case of regional model, we had used a "pre-load" method, and over-assignment to toll roads was a problem which we could dampen with "calibrating" the value-of-time / perception weights for the base year.
In our state(province)-wide model, we have not adopted the pre-load method, although it should be just as applicable. In this situation, we're faced with a bigger challenge as we've adopted a planning-level dynamic assignment of multi-classes with generalised cost (for toll roads), of having to deal with different classes of demand, even in the passenger vehicle trip assignment, i.e., urban vs inter-urban demand.
In short, we don't have any solution to offer, but will be eager to see what other agencies are doing.
Regards,

Sundar Damodaran, P.Eng.
Team Leader | Systems Analysis & Forecasting
System Planning Branch |Integrated Policy & Planning Division | MTO
416-585-7304 (voice mail) |437-993-1595(cell)

________________________________
From: alexander.o.bettinardi=odot.state.or.us@mg.tmip.org on behalf of Alex Bettinardi
Sent: Saturday, January 2, 2021 10:00:23 AM
To: TMIP
Subject: [TMIP] Improving Truck Assignment

CAUTION -- EXTERNAL E-MAIL - Do not click links or open attachments unless you recognize the sender.

I have a transportation assignment question for the wild world as 2020 comes to a close.

I'm wondering what ideas / treatments have been applied to try and keep trucks on major freight corridors as opposed to picking an optimal path by simply time, distance, and cost.

As background, this is for Oregon's Statewide model, where we assign freight and passenger traffic separately. Over the years we have tested various weights of distance versus time, but neither really solves the issue. In either case (distance priority or time priority) there can be odd network configurations where a shorter time or distance path is through a neighborhood versus on the freeway. So if we shift the priority some sections look better as far as attracting the right amount of freight traffic and some sections look worse. Shift it again, and different areas look better and different areas look worse. To be clear, we block links where trucks are not allowed to travel (and this helps), but there are many links where trucks can travel to reach their shipping destination but they are not primary routes, they are just side routes to finalize the last mile of delivery. But assignment doesn't know that and uses them at the first instance of those routes becoming optimal.

We recently had a peer state give us an interesting idea - pre-load the assignment with trucks first, so that trucks make a longer trip length decision without knowledge or the easy ability to shift based on local congestion changes. This idea was very interesting, but after looking at it more, it seems like that would be problematic for analysis in Oregon. Specifically as Oregon continues to move forward with tolling analysis. If trucks see a congestion free network first, they would be the first to jump ship at a policy made bottleneck like a toll. So where some might expect trucks to be the last vehicles to try and avoid the toll; if they are assigned first, they would likely be the first (most likely) vehicles to re-route around the toll.

My current plan/thought is to tag links with national freight highway designation. Any link tagged as a freight route would see the normal link impedance in assignment (1x). All links not tagged as a freight route would have impedance multiplied up by something like 2x (would need to play with the multiplier). This would allow trucks to dodge tolls on other major freight routes, but it would put a high penalty on dodging a toll through a small commercial street. It would also push the trucks to get on freight corridors as fast as possible and stay there for as long as possible - which I believe would be pretty close to reality.

A long question for the end of a long year. Please share your experience and thoughts with me. Many thanks

Alex Bettinardi
--
Full post: https://tmip.org/content/improving-truck-assignment
Manage my subscriptions: https://tmip.org/mailinglist
Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13526

vajucywym

Normal 0 false false false EN-US X-NONE X-NONE
Alex,
By odd coincidence, I was just beginning to do a write-up on a new method for truck assignment that we developed to create the official estimates of national truck flows derived from the new FAF5 data.
You have correctly observed that there are issues in the way that trucks are treated in a multi-class user equilibrium traffic assignment.  Pre-loading is actually quite popular but is not a good solution for many reasons.
For the nationwide FAF analysis for FHWA that treats only a small percentage of the overall traffic stream, a conventional equilibrium model was rejected at the outset as being neither practical nor realistic.  It wasn’t just that we didn’t want to attempt to model all traffic on a national network, it was the recognition that long haul truck traffic does not follow equilibrium routing principles that led us to an alternative approach.
Without getting too deep into the weeds, in a multi-class user equilibrium assignment, the overall solution is computed with passenger car equivalents (PCEs) and generalized costs reflecting tolls that vary by class.  Neither the class flows nor the route flows are necessarily unique, which is a most inconvenient problem for toll road analysis and truck assignment.  In multi-class equilibrium assignment, the solution is really defined by the travel times or generalized costs, and the same congested travel times or generalized costs could be achieved by total PCEs and tolls entirely from cars or from varying combinations of cars and trucks of different sizes.  The congested travel times or generalized costs (and total flows) are unique, but the class flows are not. 
Moreover, there is no reason why long haul trucks should follow routes dictated by equilibrium conditions such that all utilized routes have equal minimum costs.  There are many reasons for this.  Carriers and owner-operators make routing decisions based on many factors in addition to travel times and toll costs.  Many of these factors are latent and probably not knowable, but these could include time deadlines for shipments, intermediate stops, work rules, location of truck stops, and backhaul considerations.
For the FAF work, we chose a discrete route choice model that is applied for each O-D pair.  This is a path-based model with explicit route enumeration.  As many are aware, a k-shortest path process typically generates only minor variations on a route and not significant alternative routes.  To address this, a different route enumeration process was developed that proved to be surprisingly effective.  The routes enumerated are all explicit, viewable on a map, and routes can be edited, deleted, and added.  They also can be validated using appropriate data. When compared against actual truck routes from ATRI’s truck GPS data, the process captured the utilized routes to a surprising degree.
The route choice model is a path-size logit model that is based on congested travel times and accounts for route overlaps and tolls. The assigned flows and the select zone/link analyses are all unique and can be flexibly queried.  In theory, other terms such as facility classes could also be added to the utility function, but we have not tested this as yet. 
A new national network was developed by Caliper and was tagged with data from HPMS and NPMRDS.  Data from the latter source were used to obtain congested travel times for included links.  The assignment model was tested with data from FAF4 and appeared to work well.  This work was presented in summary form at TRB last year.
We are now applying this methodology to the new FAF5 data and making some enhancements along the way.  I believe that this assignment method can be integrated with an overall statewide traffic model that follows equilibrium principles for most traffic and that it should work effectively.
I would be happy to supply further details if you wish.
Howard

From: Alex Bettinardi
To: TMIP
Sent: 1/2/2021 10:01 AM
Subject: [TMIP] Improving Truck Assignment

I have a transportation assignment question for the wild world as 2020 comes to a close.
I'm wondering what ideas / treatments have been applied to try and keep trucks on major freight corridors as opposed to picking an optimal path by simply time, distance, and cost.
As background, this is for Oregon's Statewide model, where we assign freight and passenger traffic separately.  Over the years we have tested various weights of distance versus time, but neither really solves the issue.  In either case (distance priority or time priority) there can be odd network configurations where a shorter time or distance path is through a neighborhood versus on the freeway.  So if we shift the priority some sections look better as far as attracting the right amount of freight traffic and some sections look worse.  Shift it again, and different areas look better and different areas look worse.  To be clear, we block links where trucks are not allowed to travel (and this helps), but there are many links where trucks can travel to reach their shipping destination but they are not primary routes, they are just side routes to finalize the last mile of delivery.  But assignment doesn't know that and uses them at the first instance of those routes becoming optimal.
We recently had a peer state give us an interesting idea - pre-load the assignment with trucks first, so that trucks make a longer trip length decision without knowledge or the easy ability to shift based on local congestion changes.  This idea was very interesting, but after looking at it more, it seems like that would be problematic for analysis in Oregon.  Specifically as Oregon continues to move forward with tolling analysis.  If trucks see a congestion free network first, they would be the first to jump ship at a policy made bottleneck like a toll.  So where some might expect trucks to be the last vehicles to try and avoid the toll; if they are assigned first, they would likely be the first (most likely) vehicles to re-route around the toll.
My current plan/thought is to tag links with national freight highway designation.  Any link tagged as a freight route would see the normal link impedance in assignment (1x).  All links not tagged as a freight route would have impedance multiplied up by something like 2x (would need to play with the multiplier).  This would allow trucks to dodge tolls on other major freight routes, but it would put a high penalty on dodging a toll through a small commercial street.  It would also push the trucks to get on freight corridors as fast as possible and stay there for as long as possible - which I believe would be pretty close to reality.  
A long question for the end of a long year.  Please share your experience and thoughts with me.  Many thanks
Alex Bettinardi --
Full post: https://tmip.org/content/improving-truck-assignment
Manage my subscriptions: https://tmip.org/mailinglist
Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13526

covomopez

There may be a bit more to it than a single cost function for all trucks, especially if  fuel consumption and emissions outputs are of interest. First there is the split between articulated and rigid vehicles. Then there is the range of different size categories. Finally there is the split between loaded and unloaded vehicles. Much of this can be gleaned from generic published sources, combined with detailed local vehicles surveys.
Geoff Hyman 

-----Original Message-----
From: Pedro Camargo
To: TMIP
Sent: Mon, Jan 4, 2021 4:38 pm
Subject: Re: [TMIP] Improving Truck Assignment

Alex,
It looks like you have already explored the problem a bit, but I
will add my 2c.First of all, pre-loading truck trips sounds like a terrible idea and
reminds me of how things used to be done a couple of decades ago, and your
concern with policy testing is right on the money.I would submit that going towards a more sophisticated generalized cost
function for trucks is definitely the best path, but I would suggest taking
a somewhat more sophisticated approach (assuming you have Truck GPS data),
and proceed to estimate a route choice model exploring all the variables
you may have (distance, free-flow time, congested travel time, road
hierarchy, legal restrictions, etc) in order to build or generalized cost
function.Depending on your network and the OD pairs in question, treating for path
overlap might be needed and you would most likely have to adjust the scale
of the final generalized cost function when you bring it back to your
multi-class assignment, and while the latter is easy to deal with, the
former might be a little more laborious to implement.You can also calibrate a cost function by trial-and-error, but that is both
less robust and a lot less fun. If you decide to go that route, I would
recommend looking into the truck route choice literature, as you might be
able to borrow some parameters from which you can start your exploration.Again, just my 2c.Cheers,
PedroOn Sun, Jan 3, 2021 at 1:04 AM Alex Bettinardi <
alexander.o.bettinardi@odot.state.or.us> wrote:> I have a transportation assignment question for the wild world as 2020
> comes to a close.
>
> I'm wondering what ideas / treatments have been applied to try and keep
> trucks on major freight corridors as opposed to picking an optimal path by
> simply time, distance, and cost.
>
> As background, this is for Oregon's Statewide model, where we assign
> freight and passenger traffic separately. Over the years we have tested
> various weights of distance versus time, but neither really solves the
> issue. In either case (distance priority or time priority) there can be
> odd network configurations where a shorter time or distance path is through
> a neighborhood versus on the freeway. So if we shift the priority some
> sections look better as far as attracting the right amount of freight
> traffic and some sections look worse. Shift it again, and different areas
> look better and different areas look worse. To be clear, we block links
> where trucks are not allowed to travel (and this helps), but there are many
> links where trucks can travel to reach their shipping destination but they
> are not primary routes, they are just side routes to finalize the last mile
> of delivery. But assignment doesn't know that and uses them at the first
> instance of those routes becoming optimal.
>
> We recently had a peer state give us an interesting idea - pre-load the
> assignment with trucks first, so that trucks make a longer trip length
> decision without knowledge or the easy ability to shift based on local
> congestion changes. This idea was very interesting, but after looking at
> it more, it seems like that would be problematic for analysis in Oregon.
> Specifically as Oregon continues to move forward with tolling analysis. If
> trucks see a congestion free network first, they would be the first to jump
> ship at a policy made bottleneck like a toll. So where some might expect
> trucks to be the last vehicles to try and avoid the toll; if they are
> assigned first, they would likely be the first (most likely) vehicles to
> re-route around the toll.
>
> My current plan/thought is to tag links with national freight highway
> designation. Any link tagged as a freight route would see the normal link
> impedance in assignment (1x). All links not tagged as a freight route
> would have impedance multiplied up by something like 2x (would need to play
> with the multiplier). This would allow trucks to dodge tolls on other
> major freight routes, but it would put a high penalty on dodging a toll
> through a small commercial street. It would also push the trucks to get on
> freight corridors as fast as possible and stay there for as long as
> possible - which I believe would be pretty close to reality.
>
> A long question for the end of a long year. Please share your experience
> and thoughts with me. Many thanks
>
> Alex Bettinardi
> --
> Full post: https://tmip.org/content/improving-truck-assignment
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13526
>--
Full post: https://tmip.org/content/improving-truck-assignment
Manage my subscriptions: https://tmip.org/mailinglist
Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13526

ajhorowitz

Obtaining good truck assignments is not an exact science, but there are better and worse ways of doing things. 

Beyond my writings on the subject, mostly describing what others have done, I once led the development the MVFC Freight Model, which gave me a deep appreciation for the difficulties involved.  I would first like to briefly describe the MVFC Freight Model, then say a little about the traffic assignment step. 

This model covered a 10-state area.  It microsimulated freight demand at the shipment level.  Shipments were produced and consumed by establishments, not zones.  Most of these shipments were carried by trucks.  Truck shipments were organized into tours, going from a geographically exact point of production to a geographically exact point of consumption, perhaps spending time at one or more transshipment points.  Traffic assignments were dynamic and multi-class, spanning many days at 1-hour intervals.  Different commodities were placed in different truck classes due to their unique shipment costs.  There was no attempt at equilibrium because passenger cars were not loaded to the network.  Routing was sensitive to costs, and all toll roads and toll bridges in the region were identified.  Rest periods were introduced along the way for the longest hauls. 

So I learned a few things and had some beliefs confirmed: 

  • It is really important to get correct truck values of time.  Many values of time reported in the literature are not accurate. 
  • There is no substitute for a high-level of spatial precision. 
  • Many highway networks, particularly as to local streets, are not very well suited for getting the accurate impedances for trucks.  Tweaks may be necessary. 
  • Many trucks travel directly from the point of production to the point of consumption, but many others do not. 
  • A “truck” model is a big simplification of a “freight” model.  “Truck” models can be quite useful, but their limitations need to be recognized. 

I should mention that the MVFC Freight Model did not use exact establishments outside the 10-state area.  The remaining 38 contiguous states were organized into FAF zone-sized “super-establishments”.  The model simulated all truck traffic in the 48 contiguous US for the selected commodities. 

To answer a separate question, for regional models I would recommend a multi-class, equilibrium traffic assignment, but honestly, I am not sure it makes a heck of a lot of difference. 

Alan

 

Alex Bettinardi

There has been so much great feedback on this question from the TMIP community – thank you to all.
One quick follow-up question to the bullet points in your response. It sounds like you may have developed a sense for the “correct” value of truck/freight value of time. This is what we are currently using ($39/hr $2010 dollars): https://github.com/tlumip/tlumip/wiki/Traffic-Assignment#link-impedance

Extra lit review - https://github.com/tlumip/tlumip/wiki/CT#observations-on-truck-value-of-...

Would you be willing to share your experience on a good number for freight value of time?

Alex Bettinardi, P.E.
503.986.4104
http://www.oregon.gov/ODOT/Planning/Pages/default.aspx

From: HOROWITZ=uwm.edu@mg.tmip.org On Behalf Of ajhorowitz
Sent: Wednesday, January 6, 2021 7:47 AM
To: TMIP
Subject: Re: [TMIP] Improving Truck Assignment

This message was sent from outside the organization. Treat attachments, links and requests with caution. Be conscious of the information you share if you respond.

Obtaining good truck assignments is not an exact science, but there are better and worse ways of doing things.

Beyond my writings on the subject, mostly describing what others have done, I once led the development the MVFC Freight Model, which gave me a deep appreciation for the difficulties involved. I would first like to briefly describe the MVFC Freight Model, then say a little about the traffic assignment step.

This model covered a 10-state area. It microsimulated freight demand at the shipment level. Shipments were produced and consumed by establishments, not zones. Most of these shipments were carried by trucks. Truck shipments were organized into tours, going from a geographically exact point of production to a geographically exact point of consumption, perhaps spending time at one or more transshipment points. Traffic assignments were dynamic and multi-class, spanning many days at 1-hour intervals. Different commodities were placed in different truck classes due to their unique shipment costs. There was no attempt at equilibrium because passenger cars were not loaded to the network. Routing was sensitive to costs, and all toll roads and toll bridges in the region were identified. Rest periods were introduced along the way for the longest hauls.

So I learned a few things and had some beliefs confirmed:

* It is really important to get correct truck values of time. Many values of time reported in the literature are not accurate.
* There is no substitute for a high-level of spatial precision.
* Many highway networks, particularly as to local streets, are not very well suited for getting the accurate impedances for trucks. Tweaks may be necessary.
* Many trucks travel directly from the point of production to the point of consumption, but many others do not.
* A “truck” model is a big simplification of a “freight” model. “Truck” models can be quite useful, but their limitations need to be recognized.

I should mention that the MVFC Freight Model did not use exact establishments outside the 10-state area. The remaining 38 contiguous states were organized into FAF zone-sized “super-establishments”. The model simulated all truck traffic in the 48 contiguous US for the selected commodities.

To answer a separate question, for regional models I would recommend a multi-class, equilibrium traffic assignment, but honestly, I am not sure it makes a heck of a lot of difference.

Alan

--
Full post: https://tmip.org/content/improving-truck-assignment
Manage my subscriptions: https://tmip.org/mailinglist
Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13526

Alex Bettinardi

I wanted to quickly share this summary wiki page I have been building with all the great feedback on this question. If you happen to review it and feel I am missing any guidance please let me know and I will add any missing pieces in.

https://github.com/tlumip/tlumip/wiki/Truck-Assignment-Record-of-Discussion

Thank you again

Alex Bettinardi, P.E.
503.986.4104
http://www.oregon.gov/ODOT/Planning/Pages/default.aspx