Replace the very basic abstraction (in the form of “cargo units”) with an actual and much more realistic model of real-world cargo ops. Again, this is a concept item and is about figuring out what such a feature (or rather group of features) could look like.
It’ll allow more airlines to coexist in the same area
Airports that otherwise have bad cargo demand will intentionally get cargo service, and not only just as a side effect of having available unused takeoff/landing weight
My concerns:
Cargo demand is pretty scarce in most airports, specially outside of Europe, east asia or North America, So I am afraid that
will bring this scarce demand even further down
2. (more relevant towards airlines that operate flights with way fewer seats than the aircraft’s maximum), this might hurt airlines that wish to adopt a very luxury-oriented identity, or airlines that are very cargo-oriented if not cargonly
I figure once we address cargo properly, we need to forget about how cargo right now, including where there is demand or isn’t. The system will likely work in a completely different, hopefully much more realistic way.
You might want to also add the urgency of the cargo to the list: There are few but high-priced urgent cargo, think of the FedEx overnight or UPS Next Day Air, in which they will book the fastest route possible and pay a significant amount, which might even include using short haul cargo connections, compared to regular airmail which is low price and might even select medium and long haul road network on both ends.
You might even want to extend the maximum transfer time for cargo because you really should be able to have cargo held at a transfer airport to wait for the next flight on the next day if no other option is available.
I hope that the airframe is also linked to cargo type, which means the belly cargo will be most restricted (nothing dangerous and flammable etc), and you have planes like An124 that can carry special oversized cargo that other planes can’t. The current cargo system makes An124 pretty much useless
I kind of disagree with this statement. You might lack urgent high-value cargo, but you would definitely have things like airmail or special cargo such as fuel, livestock etc, The current cargo demand system is actually incorrect because there are a lot of zero cargo demand airports, which doesn’t make much sense: as long as you have passenger demand, you really should have some cargo demand as people would generate cargo. Sometimes, you might even have more special cargo demand as there are places you can only transport goods using flights due to a lack of ground network.
I’m not against dangerous goods on dedicated freight. Belley cargo are the ones on passenger flights. Actually what i suggested means you can only carry dangerous goods on dedicated cargo flights
I think the change to the cargo area is great. There are many regions in the world where a lot of mail and parcels are transported daily by smaller aircraft. At the same time, a level classification according to pure annual transport weight does not reflect everything, as many light parcels are also transported, especially in the parcel/mail/express sector, but together they take up a lot of volume. I would also hope that the change would make it worthwhile to transport freight in regional transport.
Mail and parcels are the one thing for which I have absolutely zero idea of how I could implement it.
Pax and “regular” cargo can be modelled as individual travel requests (disclaimer: at least I hope I can pull it off by cutting a few corners here and there) because there are comparatively few of them. In a remote Canadian village, half a dozen pax might fly on a Caravan and that’s manageable. But that half cubic meter of mail and parcels is in fact a bundle of possibly hundreds of tiny “shipments” to as many destinations…there’s just no way I could simulate that. The only way I can think of that might work is by “faking it” such that an imaginary postal service books mail/parcel capacity in bulk to a connected larger destination. But I haven’t given that a lot of thought yet…might end up being its own item on the roadmap.
I think some artistic license is the only way to go in a world where mail & deliveries are not just always different from day to day but constantly evolving and growing.
I wouldn’t envy the head scratching behind this one XD
At least nowadays, much of mail/parcels should be e-commerce goods, either directly from seller of Amazon etc to customers, or from suppliers via like Alibaba to shops, I believe. There are no need to simulate individual units of them.
Also, if it is desired to improve demand distribution like this, I think it is worth checking out development blog of a game called “NIMBY Rails” on steam, as its developer in recent years have recreated the passenger distribution system in that game where passengers are distributed toward players-created rail stations based on real life openstreetmaps data, and from my memory its development blog documented quite a bit of things they noted or thought of or paid attention to in the process, thus I think they are worth reading.
There sort of is. Let’s say we have a remote airport with a daily demand of 20 pax and 1 cargo unit (just as an example). In current-gen AS, there’d be a fixed table of “destinations” for this airport, which would typically boil down to around 20 destinations for pax (1 pax per day and destination) and one destination per cargo unit. These destinations would be based on our relative demand data (details on how they are picked doesn’t matter here). In next-gen AS, individual travel requests would be generated and the concrete destinations might vary from day to day, but the underlying principle remains the same. Which also means that one cargo unit would just be 100kg of cargo going to one destination every day (the same one in current-gen, a different one in next-gen).
This means that, if you have a very good network with lots of connections to almost anywhere, you probably won’t notice that this one cargo unit goes to different places every day (or always to the same destination, which you happen to connect to). But if you don’t, you either capture 0% of the traffic in the old system, or either 0% or 100% in the new system, depending on the “destination of the day”.
But for mail and parcels, that’s not how it works, right? A cargo unit made up of 100kg of mail and parcels would actually go to hundreds of destinations. Even if your network was small, you’d always expect some mail/parcel freight from a small/remote destination as it gets distributed at some sort of “mail hub”.
Since mail is typically handled by postal services, we could of course “pre-compute” that demand and just assume mail hubs in certain locations, with player airlines just hauling the (then) point-2-point traffic for the postal services. But parcels are typically transported by integrators like UPS, DHL, FedEx etc. who, as the name suggests, integrate all of the steps from door to door in-house. That sort of thing we can hardly simulate in AS…or we’d need a completely different approach for it compared to our current system.
Thanks for the heads-up! Haven’t checked it out yet (but I will), but from what you describe it sounds very similar (in principle) to what we did/do for the ASTD and individual travels request generation.
First of all, even for courier service providers like UPS, DHL, Fedex, they aren’t going to directly load a document collected directly onto airplane and then directly from the airplane to destination. For example see below one thing I recently sent/received via Fedex. The object was shipped from Yokohama, which Fedex picked up and first put it at its collection facility at Tsurumi, and then they send the package from Tsurumi to their own facility at Narita airport probably for loading into container or something. Then from Narita airport via Seoul Incheon it arrived at Chep Lap Kok International at Hong Kong. And at there it reached Fedex’s sorting facility at the airport, which then get sorted and distributed to local Fedex warehouse in elsewhere Hong Kong, before finally it get delivered from that to my home.
Second, even though it is possible for an air mail/courier item to be merely an letter, I do not think airlines in real world due with a single letter directly, and instead airlines deal with pallets or ULD containers given to them by shipment company. And if let say if there’re tons of air mails from Wenzhou of China many of them heading to Midwest US but one of them is for Bahama, I don’t think it would be the airlines who would break down the pallet/container at Chicago (or Pittsburg or) and then repack them into individual midwestern usa airport or Bahama. Instead it would be shipment companies who at their warehouse responsible for sorting and distributing them into new pallets/containers from the midwest city to Bahama/etc
Of course I don’t think airlines load individual letters. But overall demand is based on individual letters. In my “remote airport example” above, there would be one cargo unit, container, mailbag or whatever per day. But that overall unit would consist of hundreds of individual shipments. We obviously can’t simulate things on that level, hence my ideas for potential solutions, the likeliest being that we pre-compute mail networks and then merely distribute the resulting point-to-point demand somehow (which would then be on “cargo unit” level again).
But maybe there’s simply a misunderstanding here…I am looking at this from the point of view of technical implementation, you probably aren’t
XD
