On 2007-12-29 14:31:42 +0000, The Natural Philosopher said:

Owain wrote:
The Natural Philosopher wrote:
Very good case for a nationalised subsidised postal service...

One that delivers several times a day, can deliver urgent parcels
before 9 am next day, can collect and delivery local mail same day, and
can collect from street receptacles 7 days a week ...

We tried it once, but the unions went on strike and broke it, so
everyone bought fax machines instead.


The piojnt is that fir messagiung, text, email and even fax is FAR more
reliable.

Its only GOODS that need themovement, and I query the need for several
times a day eirther.

On or two more or less guaranteed deliveries with a completely reliable
tracking system would be a start.

Let's consider this,because I did a few years back.

Every packet, or box, has a barcode on it, that as a minimum, the
target address (postcode, and house, unique) and a unique packet ID.

Automatic scanners, or manual scaners if the autos do not work, scan
each parcel that is recieved, at any given point in its journey.

The scanners relay the position of te parcel back to a central series
of computers that propagate the information between them.

Each scanning station runs the equaivalent of a dynamic routing
protocol. Vehicles travel between scanner stations (Routers) with the
packets, and load up with the packets to the gunwales, on a 24x7 basis.

When they arrive the packets are rescanned, and re-rtouted for the 'next hop'

If a packet is clocked IN to a vehicle, but not OUT, it is flagged as
an alarm, and the carrier notified.

At ANY stage of this the last known location of a packet could be established.,

You could arrange for e-mail notification when it gets to the nearest
routing station, and pick it up, or have the local lower QOS postal
service deliver it.

At the pickup point, it gets 'removed' from the 'active' system.

Now to may way of thinking, you have a nationalised central database
system for this, or at least a national standard for data interchange.
This is 'the parcel internet'

Each routing station could be private, run for profit, and paid on a
per parcel basis. This is the 'parcel ISP'

Each carrier that picks up parcels could be anyone, froma train to a
man with a van, paid on a per packet basis.

This is the 'parcel carrier network'


As with broadband, the last mile is the biggest problem. However again
a local man with a van, is probably as good as it gets.

As with the internet, who you pay and how you pay them is a
variable..you would probably have a standard rate and pay the local
carrier. HE gets charged for onward delivery ...if you make a standard
rate for a kilogram packet mile, or some such, then the system will
automagically sort out what it will cost, and do least cost routing, or
however you arrange it..

No GUARANTEES, but as with the internet, a 99% probability of 10x
better service than a 100% guaranteed service is usually good enough.

And the packets are totally trackable through every single station..and
if things go wrong, they can be unloaded at any station and resinserted
into the system and will still get forwarded along to the right 'next
hop'


This is lovely, but for one thing.

Presumably there would be a nominated approximate time for each
property. That's useful and would have the advantage that the
recipient would have a reasonable time window where they know that they
will get a delivery (or not if there isn't one) and so can plan their
day.

Most couriers work by loading up their van with parcels organised so
that earliest deliveries are first off and so on and then begin picking
up at some point during the day.

However....... customers will want to buy premium delivery service.
For example, if I buy something from RS, I can choose before 0900,
before 1000 or next day, or for especially urgent things, a van/bike
arrangement.
Suppliers will be eager to accomodate them because it is a
differentiator based on service and money is attached to that in
several ways.

At that point, bang goes the dynamic routing protocol of the model and
in comes the pre-emptive forwarding, the weighted fair queuing and even
the TE of the delivery model.

Of course, one could achieve most of the desired effect, as it is done
in the case of the internet by massively over-provisioning the core and
interconnectivity bandwidth to the major places. However, that would
be self defeating for an eco-solution because the point is that that
capacity should be run at the limit and not half empty.

As it is with the internet today, the failing would be at the edges
(collection and delivery) and with connectivity to the more difficult
countries where the volume or political situation don't allow the best
service.