Commercial grade none proprietary software to drive CIS

[Joe]

Hello,

Is there any none proprietary software available that TOC's could use which would do away with there need for a Worldline / Ketech feed/contract for CIS?

I understand the feed directly from DARWIN is not good enough and doesn't capture everything. 

My understanding of the feeds to CIS is that there is always a Worldline / Ketech feed required but if all the information TOCs have themselves is fed back into DARWIN. I can't understand what Worldline / Ketech are providing that justifies the price tag?

Thanks in advance,

Joe

[Peter Hicks]

Hi Joe

My $0.02: I see your thinking here.  The CIS supplier provides more than just the software - it's the interfaces to displays at stations, recording and updating the automated announcements (or "doing a Temu" and using Amazon Polly for that inauthentic, unsettling feeling), interfaces to DCIS, support, etc.

It would absolutely be possible for the rail industry to take something like this in-house.  LICC was, I think, a British Rail product before the privatisation in the 1990s and WebLICC is a modern, updated and rearchitected version of this.  The downside is that you will always need an organisation owning the specification, building and maintaining software and providing support for it.  It'll need ownership, as any Open Source project does.

What might be interesting would be to have an open architecture for CIS and associated systems.  An open, documented and extensible protocol for showing data on displays, an API for making audio announcements, etc.  The original BR1810 protocol (of which the TD S-Class and C-Class messages are a tiny subset) had scope to interface with customer-facing displays, but I'm not at all confident this is still in widespread use.

Maybe this is something for GBRX.

Peter

[David Wheatley]

Hi Joe,

I suppose it's something I've played around with on my websites, which can trigger live announcements and display upcoming services in a departure board style.

The main point of people like Worldline and KeTech are to sandwich all the data sources together, and provide a convenient system to manage it. That with installation, maintenance, upgrades, SLAs, etc, will always command a price tag.

I would probably say that Darwin has almost all the info needed these days, and someone with enough willpower could make an open source replacement. Often, the issues left with Darwin are that the TOCs aren't providing all the info they have to the open feeds (e.g., formation info is still missing for services today, service alerts like "full and standing" on depature boards not making it into the Darwin feed, SE/GR include rolling stock info on their CIS but that's not in Darwin).

David

[Edward Hesketh]

Hi David,

I'm currently working on my own open-source project to do something somewhat similar to what you describe here - it's still in its early stages, but I'm trying to create a service that can:

• read data from Darwin Push Port
• provide a nice endpoint to query current data relating to a train, station, specific platform, etc.
• provide a subscription feed for realtime update messages relating to trains/stations/platforms/etc.
• be simple to selfhost
• by well documented
• have easy to read code
• be easy to expand with more data sources in the future

It's certainly not a CIS solution in its entirety, but should be useful for reducing the work required for others to create stuff using realtime data. I want it to be the sort of thing that people would find easy to write a live frontend for (thinking: apps, websites, boards, announcements), which currently isn't so easy due to how Darwin outputs data.

The main purpose is just to generate messages with more meaning and context behind them - "location y is cancelled for train x" rather than "here's a new schedule for train x", and also to provide a nice way of querying train data from Darwin on demand. ("what are the IDs of all trains departing LEEDS in the next 10 minutes" or "what is the schedule of train ID x"?)

Just saw this thread and thought you might be glad to hear there is someone crazy enough to spend 130+ hours (so far) trying to make Darwin a little easier to work with.

It's on my github under headblockhead/railreader if anyone is interested - it's a long long way from being complete, but it's going well enough so far that I'm willing to share it publicly.

Sorry for derailing the thread a little, I'm just quite passionate about my project and wanted people to know that I was working on something related.

- EH