[transferred from OEO] Add `modal split`

[lumi321]

Description of the issue

In many decriptions of the transport sector a modal split (also known as modal share or modal choice) is used. This describes the percentage of a transport modality compared to all the modalities considered.

Ideas of solution

There are several possibilities for implementation.
Possibility one: modal split as subclass of fraction value and axioms 'is about' some 'vehicle' and 'has unit' some 'percent'

Possibility two: like above, but with modal split share instead of just modal split. In addition, modal split as subclass of data item with axiom 'has part' some 'modal split share'.

The latter possibility distinguishes between the "whole" modal split and single parts. In contrast, the first option simply states the share of a certain modality without having a connection to something overarching.

There is also a third option where modal split is not even a fraction value but simply a data item that 'is about' some 'vehicle'.

In my opinion, possibility 2 is the most accurate. The main question here is if that level of detail is necessary/wanted or if a simpler solution like possibilities 1 or 3 is preferred.

Definitions depend on the choice above, but are probably variations of this:
A modal split (share) is a fraction value/data item that describes the share of a transport modality compared to all considered transport modalities.
A modal split is a data item that represents the usage of transport modalities.

Workflow checklist

• I discussed the issue with someone else than me before working on a solution
• I already read the latest version of the workflow for this repository
• The goal of this ontology is clear to me

I am aware that

• every entry in the ontology should have a definition
• classes should arise from concepts rather than from words

[l-emele]

To me, one mode of transportation (e.g. bicycles) have a modal share and the combination of all modal shares is the modal split. We can see a modal share also as an attribute of some transport process.
So what about about:

• A modal share is a fraction value/process attribute that describes the share of a transport mode.
• A modal split is a data item that contains modal shares.

General remark: We should discuss the definition first, the axiom should be derived after from the definition.

[lumi321]

Sounds good. I would prefer fraction value for modal share though. Of course, a single transport process contributes with the used modality to the modal share, but the modal share itself does not existentially depend on a single process. The way I interpret the definition of process attribute that is a requirement though.

[l-emele]

Okay, I can agree to this thought.

[lumi321]

Summarisation:
New classes:
modal split: A modal split is a data item that contains modal shares.
modal share: A modal share is a fraction value that describes the share of a transport mode.
New axioms:
'modal split' 'has part' some 'modal share'
'modal share' 'is about' some 'vehicle'
'modal share' 'has unit' some 'percent'

Do we also agree on the bottom two axioms?

[l-emele]

I am unsure about the axiom 'modal share' 'is about' some 'vehicle'. Following the definition, we need here more something like 'modal split' 'is about' some 'transport mode'. If have no immediate idea how to define a transport mode. But it has subclasses like walking, public transport, cycling or driving a car. So it is a split of processes (in which vehicle can be involved, but don't necessarily need to as in the case of walking).

[lumi321]

As far my research tells me, the subclasses of transport mode can be very different depending on the subject of interest. There is a section on the german wikipedia page about modal split that shows what I mean. So, defining such subclasses could be complex. For now, I don't think that is necessary.

• Picking up your thought from earlier to add a process attribute, the transport mode could be one. At least, if we argue that every transport process has its own transport mode.
• Looking at the other classes from the BFO, it cannot be a specifically dependent continuant. In that case, transport mode can only reasonably be connected to vehicle ('vehicle' 'has quality' some 'transport mode'), but that would exclude walking again.
• I believe it is also fair to say that transport mode is not an independent continuant.
• This leaves the generically dependent continuant or more precisely the information content entity, which then would be about some transport process. (Which may or may not include a vehicle.)
• That leads my thoughts towards another, more condensed possibility: 'modal share' 'is about' some 'transport' (process), for which we would need to change the definition of modal share to ... describes share of a transport mode process.

I can see pros and cons for all options, so I don't have a preference here.

[lumi321]

If there is no preference, I opt for the last option, changing the definition and leaving transport mode out.
From my view this adapted definition is not wrong, but limited in its expressiveness compared to a solution with transport mode. Since this expressivness may be desired in the future, I would add an editor note saying: The definition may be altered to "... transport mode" instead of "... transport process" once a definition for transport mode has been agreed on.

[lumi321]

I have one more suggestion for transport mode:
A transport mode is a process attribute that describes a transport process. It is used to distinguish between transport processes by looking at certain characteristics such as propulsion type, environment, purpose, ownership, etc.
Axiom: 'process attribute of' some transport

If that one doesn't suffice, I would continue with the temporary solution from my last comment.

[lumi321]

@OpenEnergyPlatform/oeo-domain-expert-energy-modelling any thoughts on the definition of transport mode?

[stap-m]

If have no immediate idea how to define a transport mode. But it has subclasses like walking, public transport, cycling or driving a car.

If these ⬆️ are examples for transport mode, it should be classified as process. Otherwise, if classified as process attribute, what would examples of transport mode look like @lumi321 ?

If there is no preference, I opt for the last option, changing the definition and leaving transport mode out.

transport mode is rather central for the current definition of transport split. Thus, if the latter stays as is, I'd recommend to define transport mode as well.

[lumi321]

I wouldn't define transport mode itself as a transport process. At least to me, a definition like "A transport mode is a transport process that ..." doesn't make a lot of sense. However, the examples walking, cycling, driving are processes indeed.
So, if all transport modes are included as transport processes, the modes would become unnecessary. Then we could take one of my previous suggestions: "A modal share is a fraction value that describes the share of a transport mode process."

If transport mode is classified as process attribute, examples would be by foot, by bicycle, by car, public, private. I have to say though that I am not convinced of this anymore.

---

Perhaps it's a good idea to discuss what transport mode actually should capture. My motivation was to find a way to represent distinctions or categories in the transport sector, aka transport mode/mode of transport. In german usually "Verkehrsträger" or "Verkehrsmittel". Examples of such categories are in the images below:
¹

²

³

As can be seen, there are many different categories and some of them, like in the last picture MIV-Fahrer/private motorised transport driver and MIV-Mitfahrer/private motorised transport co-driver, are everything but handy. This one is also a good example, where I don't know how it could suitably be defined as a transport process.
So, for me transport mode is a category of transport processes. Thinking about that, I wonder, if information content entity wouldn't be the better parent class, i.e. A transport mode is an information content entity that describes a category of transport processes.

Footnotes

1. Image taken from: Energiewirtschaftliches Institut an der Universität zu Köln (EWI) (2021). dena-Leitstudie Aufbruch Klimaneutralität. Klimaneutralität 2045 - Transformation der Verbrauchssektoren und des Energiesystems. Herausgegeben von der Deutschen Energie - Agentur GmbH (dena). ↩

2. Image taken from: Prognos, Öko-Institut, Wuppertal-Institut (2021): Klimaneutrales Deutschland 2045. Wie Deutschland seine Klima-ziele schon vor 2050 erreichen kann. Langfassung im Auftrag von Stiftung Klimaneutralität, Agora Energiewende und Agora Verkehrswende ↩

3. Image taken from: infas, DLR, IVT und infas 360 (2018): Mobilität in Deutschland (im Auftrag des BMWI) ↩

[l-emele]

Maybe we should try to find a definition for transport mode without a classification into an parent class first, so something:
A transport mode is an X that ...

When we've settled on a definition, we probably find the classification. So let's start with the last proposal: A transport mode is a X that describes a category of a transport process.

What exactly does category mean here? Something like a type? And how does this category distinct from a simple subclass like public transport?

[lumi321]

Well, I would say a category here means a subset, which is the very definition of a subclass. As I said earlier, I wouldn't mind if we do it that way.
The only "problem" I see with defining subclasses of transport is that some subclasses would be tricky to define, e.g. micromobility or the distinction driver/co-driver in private motorised transport like in the example above.

[lumi321]

I have compiled a list with terms that provide a good coverage of the transport modes used in modeling. I would like to have them included either as subclasses of transport mode itself or as subclasses of transport. The wording of the terms may be changed for better distinction.

The list is based on common distinctions of transport modes:
passenger or freight
public or private
road, rail, air, or water
national or international
small, medium, or big/heavy (truck only)
short, medium or long distance
Not all combinations are used, though, and there are some additions like micromobility with walking and cycling, distinction between driver, co-driver, and car/ride sharing, and pipeline transport. EDIT: Added local public transport to the list because a simple conjunction of bus and rail modes may not suffice to represent the local public transport.

• micromobility
  • walking
  • cycling
• passenger
  • private
    • driver
    • co-driver
    • car/ride sharing
  • public
    • local public transport
    • road
      • city bus
      • regional bus
      • coach/long distance bus
    • rail
      • underground/subway
      • regional train
      • long distance train
    • air/avation
      • national
      • international
    • water/ship
• freight
  • road
    • small truck
    • intermediate truck
    • big/heavy truck
  • rail
  • air/aviation
    • national
    • international
  • water/ship
    • national
    • international
  • pipeline

Regarding the definition of transport mode:
The function of a transport mode is to describe a type of transport, and it is used to distinguish between different types of transport. That is also what a subclass of transport does/is used for. So, from that I would say that A transport mode is a ??? that describes categories/types/subclasses of transport. Or maybe even shorter: A transport mode is a ??? that describes transport.

[lumi321]

I will add this issue to the list for the next dev meeting.