Eastern Railroad Discussion > "Predictive Analytics" and the railroads

Yesterday, CSX chief Jim Foote mentioned "we need to become more like UPS" in the way we do business.  Today's Wall Street Journal has an article on how UPS is using technology to predict bottlenecks and threats to its operations.  Considering UPS is already a leader in logistics, this is fascinating reading.

Practices like re-routing some cargo days in advance of predicted weather events, adjusting staffing in advance of incoming cargo surges or delays in the chain.  With railroads already operating on minimal extra boards, the ability to "call in" extra staff is restricted.  And, you can go from sixth out to next out in a flash, or find the hot shot you thought you'd get is now a slow drag.  Or the expected six hours of time and track suddenly constricts to a lot less.  Now, everyone will be able to see the same info, in the same time.“Today, we use data extensively to plan, but the more real-time data we can get on the state of a package, the better visibility we can get about any exceptions in the network, helping us generate improved plans to manage the network as a whole,” he said.Developing the tool was an ambitious feat because of how many hundreds of millions of data points needed to be consolidated into one single platform, said Ted Abebe, senior director for UPS’s advanced technology group. Until now, forecast, capacity, customer and package data was housed in different applications.The tool will be tested in the coming weeks and is expected to be available to UPS employees by the end of the year via a smartphone, desktop and tablet application.(subscription site, but the info should be available with a google search later today)

https://blogs.wsj.com/cio/2018/07/17/ups-it-chief-says-ambitious-analytics-project-will-improve-logistics-planning/?mod=hp_minor_pos5

We used to call a lot of that "common sense".

Out

It would have to be very flexible in predictions, as in how does the entire picture change when there is a delay such as a derailment, grade crossing accident or the motive power fails. What if a crewman is simply late for work? You can imagine in your head how the entire system would have to recalibrate and reset itself, nearly constantly. Still, there is a lot of room for improvement. As an Engineer, it never made much sense to haul butt especially in a loaded coal train, mercilessly guzzling fuel and pounding the infrastructure at 50 mph, then go into a siding for an hour or two to meet another train. It seemed if that could be predicted and reported by the Dispatcher, that a lower speed with less fuel use, less wear and tear on equipment, less risk of derailment and grade crossing accident would be vastly preferable. But it seemed the emphasis was simply to get out of the way in a hurry, with the attendant extra expense of doing so accepted. === === = === JLH

callum_out Wrote:
-------------------------------------------------------
> We used to call a lot of that "common sense".
>
> Out

Agreed.. and here is the great blues player Kenny Neal talking about 'Plan old Common Sense'

https://www.youtube.com/watch?v=xo8EwQIJb_g

Lackawanna484 Wrote:
-------------------------------------------------------
> Yesterday, CSX chief Jim Foote mentioned "we need
> to become more like UPS" in the way we do
> business.  Today's Wall Street Journal has an
> article on how UPS is using technology to predict
> bottlenecks and threats to its operations. 
> Considering UPS is already a leader in logistics,
> this is fascinating reading.
>
> Practices like re-routing some cargo days in
> advance of predicted weather events, adjusting
> staffing in advance of incoming cargo surges or
> delays in the chain.  With railroads already
> operating on minimal extra boards, the ability to
> "call in" extra staff is restricted.  And, you
> can go from sixth out to next out in a flash, or
> find the hot shot you thought you'd get is now a
> slow drag.  Or the expected six hours of time and
> track suddenly constricts to a lot less.  Now,
> everyone will be able to see the same info, in the
> same time.“Today, we use data extensively to
> plan, but the more real-time data we can get on
> the state of a package, the better visibility we
> can get about any exceptions in the network,
> helping us generate improved plans to manage the
> network as a whole,” he said.Developing the tool
> was an ambitious feat because of how many hundreds
> of millions of data points needed to be
> consolidated into one single platform, said Ted
> Abebe, senior director for UPS’s advanced
> technology group. Until now, forecast, capacity,
> customer and package data was housed in different
> applications.The tool will be tested in the coming
> weeks and is expected to be available to UPS
> employees by the end of the year via a smartphone,
> desktop and tablet application.(subscription site,
> but the info should be available with a google
> search later today)

Sorry, I don't see this as quite the breakthrough being touted. My uncle, a train master back in the 80's and early 90's, was using this same sort of data in this same sort of way to watch his cars coming into Selkirk and plan out his terminal work in NY and Boston days in advance of their actual arrival and request prioritization of movements by Selkirk.  The Big Data has been there for close to 40 years. If almost no one was using it to advance plan their work or manage network fluidity, what does that say about the vision of the company to use what was already at its disposal to improve service, customer satisfaction, and efficiency?  It isn't as if we just invented networked computing and databases - those came into being in the 1970's.  Maybe running operations by drinking from a firehose and having a daily panic is not an actual way to run a railroad.  Maybe mass efficiency devices like hump yards seem superfluous and costly to some because they don't understand how to use them to simplify their work.  Maybe it seems like too much effort to capture single carload freight movements because it requires just this sort of effort to even be in the business.  Anyone can run a 150 car train of coal downhill and look like a financial genius.  Apparently almost no one can handle thinking about 10,000 discrete car movements.

The average person is not taught to think creatively any more; only to react.

Lackawanna484 Wrote:
-------------------------------------------------------
> Yesterday, CSX chief Jim Foote mentioned "we need
> to become more like UPS" in the way we do
> business.  Today's Wall Street Journal has an
> article on how UPS is using technology to predict
> bottlenecks and threats to its operations. 
> Considering UPS is already a leader in logistics,
> this is fascinating reading.

I can't help but see the humor in this. I started doing an elementary version of this almost 50 years ago. I used an integrated system for data collection. It was black, had a dial for selecting data inputs, and a speaking/listening device for I/O. I used a high resolution permanent storage device for the data, a memory device of 11 x 17 inches operated in a compressed space state of 8 1/2 x 11 inches through a plication process, that could be used in read-write protected or unprotected state.

Over a couple of decades, I enhanced the system with technology, increasing the speed and accuracy of predictions. However, I was ridiculed, harassed, and threatened for doing it. Now, somebody thinks it is cool.

A railroad is not UPS. There are some similarities in some places, but in general, what applies to UPS either does not apply to a railroad, or applies in a different way. BN even admitted that they saw management of the railroad to be the same as managing a trucking company. They hired McKinzie & Company consultants (and paid a lot of money) to unsuccessfully develop the concept that couldn't work. BN had a history of doing that.

EHH's precision railroading was an extension of the Precision Execution program BN developed. Both were merely buzzwords. Precision railroading, like Precision Execution, requires something planned with precision to be executed with precision. If my plan is to drive to Chicago, executing it can't really be called precision. If my plan is to drive to Chicago and spend seven days doing it, arriving in seven days cannot be called executing with precision. If I say that I'm going to drive to Chicago leaving at 7pm Friday, via I5 to I90 to the Spaghetti Bowl interchange to Congress Parkway to Columbus Drive to the parking garage in Grant Park, determine construction zone restrictions on the route, compare the running time to the time of day of normal congestion on the route and adjust, take 1.1% of the speed limit driving time including construction zone restrictions and adjusted running time for normal congestion times and locations, make fuel stops and tie up for rest at seven hour intervals, plot those theoretical locations against potential stop locations (let's not arbitrarily stop for fuel and rest in the middle of nowhere because that's where running time says it will be), precision execution is possible. Precision execution cannot be performed without a precision plan. However, as senior BN management repeatedly admonished me, precision planning my way requires too much work. They were much more in favor of considering the activity of planning to eventually arrive in Chicago and eventually arriving in Chicago as precision execution.

I don't believe, based on what I observe on a regular basis, that anything has changed.

My guess is that CSX will follow the same path, attempting to fit the 2" diameter round peg into the 1" square hole, beating the peg into submission, and finally quietly moving on the next big thing.

TAW

 

Didn’t NS just post something like this on that big video site?

Posted from iPhone

PRR1361 Wrote:
-------------------------------------------------------
> The average person is not taught to think
> creatively any more; only to react.

...beyond the average person, the folks in charge are taught the same thing. They just pay a lot more for it.
TAW

highgreengraphics Wrote:
-------------------------------------------------------
> It would have to be very flexible in predictions,
> as in how does the entire picture change when
> there is a delay such as a derailment, grade
> crossing accident or the motive power fails.

I have written extensively in Elements of Train Dispatching Vol 2 and in Managing Railroad Transportation just how to do this. There are procedures that I used successfully for decades. It's not rocket science. Maybe if I dressed it in complexity, used a bunch of Greek symbols, and dressed it as rocket science, I would be more successful in promoting the idea,.


> What
> if a crewman is simply late for work? You can
> imagine in your head how the entire system would
> have to recalibrate and reset itself, nearly
> constantly.

BN tried such a global approach in the 90s and spent a lot of money dreaming the impossible dream.

> Still, there is a lot of room for
> improvement. As an Engineer, it never made much
> sense to haul butt especially in a loaded coal
> train, mercilessly guzzling fuel and pounding the
> infrastructure at 50 mph, then go into a siding
> for an hour or two to meet another train. It
> seemed if that could be predicted and reported by
> the Dispatcher, that a lower speed with less fuel
> use, less wear and tear on equipment, less risk of
> derailment and grade crossing accident would be
> vastly preferable.

That was my way of working for a couple of decades. It included knowing where I could send crews for coffee/food to burn time instead of sending them to outer boondocks to look at snakes and rabbits while waiting. I know a guy in Switzerland who developed a signal system to predict the required traffic flow for the single track section of the Lötschberg Base Tunnel. It's a well accepted concept in Europe, but I have been repeatedly told by the experts that European stuff doesn't work.

Add to your observation, the concept of "plenty of green" for Red Hot Smokin' Awesome Z Trains and you have a gigantic hole in the pressurized vessel, blowing cash out of it. There is no window near the hole, so the loss of cash is mere a matter of unexplained cost of doing business.

TAW



Edited 1 time(s). Last edit at 07/19/18 10:22 by TAW.

abyler Wrote:
-------------------------------------------------------
> Sorry, I don't see this as quite the breakthrough
> being touted. My uncle, a train master back in the
> 80's and early 90's, was using this same sort of
> data in this same sort of way to watch his cars
> coming into Selkirk and plan out his terminal work
> in NY and Boston days in advance of their actual
> arrival and request prioritization of movements by
> Selkirk.  The Big Data has been there for close
> to 40 years. If almost no one was using it to
> advance plan their work or manage network
> fluidity, what does that say about the vision of
> the company to use what was already at its
> disposal to improve service, customer
> satisfaction, and efficiency?  It isn't as if we
> just invented networked computing and databases -
> those came into being in the 1970's.  Maybe
> running operations by drinking from a firehose and
> having a daily panic is not an actual way to run a
> railroad.  Maybe mass efficiency devices like
> hump yards seem superfluous and costly to some
> because they don't understand how to use them to
> simplify their work.  Maybe it seems like too
> much effort to capture single carload freight
> movements because it requires just this sort of
> effort to even be in the business.  Anyone can
> run a 150 car train of coal downhill and look like
> a financial genius.  Apparently almost no one can
> handle thinking about 10,000 discrete car
> movements.

LOL, but CSX's "new" program probably comes with multi-pretty lights, diagrams, and flashy thingy's!  Well worth spending the extra millions of dollars for! <sarcasm>

Seriously though, great illustration example here of a possible reality that works.

callum_out Wrote:
-------------------------------------------------------
> We used to call a lot of that "common sense".
>
> Out

Common sense isn't so common now..... 

TAW Wrote:
-------------------------------------------------------

> A railroad is not UPS. There are some similarities
> in some places, but in general, what applies to
> UPS either does not apply to a railroad, or
> applies in a different way.

I have always thought that what UPS does is a good illustration of what happens with the railroad' carload (ie loose car) business.  UPS has to deal with a huge complex of fluctating volume (of packages) and have a logistics operation set up so that all those packages can be moved economically, yet reliably and in a time effective manner, to their customers destinations (after all, its an extremely competitive business they are in).  I think railroads could use some of what UPS has learned in handling their carload business, if they were (1) smart enough to understand what UPS is doing; (2) had the disipline enough to make the ideas work, and to continue doing it instead of dropping the ball a few weeks later; (3) have a much greater customer focus (like UPS does) in handling their business.  UPS doesn't let packages sit for days on end waiting for enough packages to load a monster size trailer, unlike railroads that let cars sit in yards for days on end so they can maximize a huge train, and then move those cars to the next yard, repeating the process.  The customer will eventually get their cars, with a variance of days and even weeks of their originally promised delivery date, which is already quite slow.
 

UPS's system is broken.  Like yesterday's package to our home that UPS delivered to someone on a different street in our neighborhood.  They don't even ring the doorbell anymore let alone make sure the package gets delivered correctly.  UPS makes it increasingly difficult to successfuly get a claim for their mistakes.  From what I can tell, UPS's technology is used primarily to make sure their drivers are run ragged and worked up to a point just shy of breaking.

But CSX's stock is up $4.56/share since yesterday's quarterly call so my observations and worldview must be wrong and CSX and their idol UPS must be right.



Edited 1 time(s). Last edit at 07/18/18 14:23 by RRTom.

I just realized after 21 and a half years, I have been through more flavor of the month initiatives than I can count....I realized also...that this one too shall pass. 

spwolfmtn Wrote:
-------------------------------------------------------
 UPS doesn't let
> packages sit for days on end waiting for enough
> packages to load a monster size trailer, unlike
> railroads that let cars sit in yards for days on
> end so they can maximize a huge train, and then
> move those cars to the next yard, repeating the
> process.  The customer will eventually get their
> cars, with a variance of days and even weeks of
> their originally promised delivery date, which is
> already quite slow.

Ding, ding, ding, we have a winner here! It's called a monopoly, and the railroads have it in spades. Lots of insightful comments here but none better than the above. Carload and unit trains all fall under the category of monopoly, so why run your business like UPS? Besides, Foote says this as they announce record profits from the way they are currently operating, so that's just lip service for the gullible. Only intermodal has any real competition. As far as using big data is concerned, it's been discussed quite extensively here on TO how railroads could do quite a bit more to provide better service, generate more high-profit carload business and such (all using big data), but there's no real willpower. Actions speak louder than words; look at EHH's model: Drop business and cut capital spending and workers to reach short-term goals, change operating practices that maximize the OR but leaves the network screaming for relief. After milking this scenario for awhile for big returns, it's dump and run and leave the mess for someone else, and many customers fuming. Classic corporate raiding, railroad-style. Trains recent articles, all unnecessarily praising EHH, admit as much if you read between the lines, and each road he left is still recovering years later, as will CSX. Having a monopoly is a wonderful thing if it's yours, but a terrible thing if it's not yours and you depend upon it. UPS had it for years before the likes of Fedex and DHL but had to change due to intense competition. Railroads will never operate even remotely close to how UPS currently does because there's no real need to. With no competition for the vast majority of your business, it's just do your business and fly under the radar enough not to get caught too badly bilking your customers with crappy service and high charges. There's no "railroad Fedex" on the horizon to worry about. It's a part of the capitalist setup to this day that's not really operating under the "free market" premise, and the government allowed it. With record profits and the customer be damned approach, do you really think they're going all in on big data for customer service?? It's like stealin' candy from a baby...

JG



Edited 2 time(s). Last edit at 07/18/18 19:18 by jgilmore.

I have a shipment to move from point "A" to point "B".  I can use rail, truck, barge, or even air freight.  How can there be a monopoly when there are so many sellers in the market?
Yes, in some cases, railroads have a "tight lock" on some traffic, but for the most part, there's lots of competition out there.  That's why 75% of all intercity freight moves by truck.  Throwing out buzzwords like monopoly do not convince anyone of your argument.

Norm

Norman Orfall
Helendale, CA
TIOGA PASS, a private railcar

aronco Wrote:
-------------------------------------------------------
> I have a shipment to move from point "A" to point
> "B".  I can use rail, truck, barge, or even air
> freight.  How can there be a monopoly when there
> are so many sellers in the market?
> Yes, in some cases, railroads have a "tight lock"
> on some traffic, but for the most part, there's
> lots of competition out there.  That's why 75% of
> all intercity freight moves by truck.  Throwing
> out buzzwords like monopoly do not convince anyone
> of your argument.
>
> Norm

For much of the traffic that moves in bulk, thus using carload "service", it is essentially a monopoly in many places (your captive by physical track access).  Much of this traffic does not have a resale value that can justify higher costs of using truck, or just plain can not be trucked effectively.  Chemical companies have been complaining about this for years.  Even when a Class one railroad sells/leases off it's "marginal" lines, they put up paper barriers to prevent the competition from getting access to those markets via the new shortline.  So in effect, there's little enthusiasm for Class one railroads to actually provide better service, though they pay a hell of a lot of lip service to it!

On the other hand, as mentioned here, intermodal is highly competitive between other railroads and trucks.  So you see railroads actually providing a competitive service (in most cases) to obtain and keep this business.

aronco Wrote:
-------------------------------------------------------
> I have a shipment to move from point "A" to point
> "B".  I can use rail, truck, barge, or even air
> freight.  How can there be a monopoly when there
> are so many sellers in the market?
> Yes, in some cases, railroads have a "tight lock"
> on some traffic, but for the most part, there's
> lots of competition out there.  That's why 75% of
> all intercity freight moves by truck.  Throwing
> out buzzwords like monopoly do not convince anyone
> of your argument.
>
> Norm

Quite right...railroads are hardly a monopoly...we know what we can haul profitably and that's what we haul. For better or worse we won't chase hauls that don't pay. The business we can't do profitably rightly goes to UPS FedEx and whoever else is out there. That's how engineer and Maintenance of Way can be Teamsters....there's very little overlap in what we haul but we all haul something. Certainly we share alot....we haul the big loads and trucks do the last mile. We haul endless quantities on a continous basis, while trucks can keep to guaranteed day and time with a truckload. Of course that truck was filled with stuff that came in a boxcar...or tanker....or hopper. 

aronco Wrote:
-------------------------------------------------------
> I have a shipment to move from point "A" to point
> "B".  I can use rail, truck, barge, or even air
> freight.  How can there be a monopoly when there
> are so many sellers in the market?

You are using the same logic that the EU uses in subsidizing freight by highway but not freight by rail. The EU policy is equal non-discriminatory access among transportation providers.

Rail freight providers must be allowed equal, non-discriminatory access to the national rail network. They must pay for the full cost of the infrastructure that they use (the railroad infrastructure is a non-subsidized toll road operating as a for-profit corporation).

Highway freight providers must be allowed equal, non-discriminatory access to the national highway network, for which they pay 20% of the cost of infrastructure that they use.

Public money may be used in the construction of new rail and highway facilities. Rail capitalization is paid back through the access fees.

Rail vs highway carriers is certainly not competitive, but by separating the modes for examination, that detail goes away unless one understands what is not being said directly.

TAW