Swindon <-> Westbury service updates and amendments, ongoing discussion - 2026

[whole diary]

[TaplowGreen]

18:37 Westbury to Swindon due 19:21
20:12 Swindon to Westbury due 20:56 will be cancelled.
This is due to a safety inspection of the track.

and

Delays to services between Newbury and Reading
Following a safety inspection of the track between Newbury and Reading the line towards Newbury will be reopened shortly.
Train services running to and from these stations are returning to normal. Disruption is expected until the end of the day.
Customer Advice
-
What has happened?
-
We have had a report of a defect with the track in the Theale area. Trains may not be able to run over the affected line until it has been inspected or may have to run at reduced speed.

Is it fair to take it that a safety inspection in Berkshire on the Berks and Hants is the case of cancellation of trains in Wiltshire on the Transwilts line?

Maybe not, but in the context of (for example) the 1636 Paddington-Plymouth having already been cancelled due to said track defect, it being a summer Friday and the weekend of Boardmasters the 1736 was undoubtedly rammed and needed to be prioritised via this diversion.

Inconvenient for customers who were planning to travel on either of those cancelled Transwilts services, but in the context of the number of customers impacted, without doubt the lesser of two evils.

[bobm]

Surprisingly, if the GWR^▸ (Great Western Railway) app is to be believed, the 17:36 appears to have seats as it approaches Castle Cary.

[Image from here is not available to guests]

The preceding 17:03 to Penzance seems to only be congested in first class.

[TaplowGreen]

Surprisingly, if the GWR^▸ (Great Western Railway) app is to be believed, the 17:36 appears to have seats as it approaches Castle Cary.

[Image from here is not available to guests]

The preceding 17:03 to Penzance seems to only be congested in first class.

I'd say "if" is doing some phenomenally heavy lifting there!

[JohnM]

Is that based on seat bookings, or actual 'live' occupancy?

[bobm]

It’s data available from IETs^▸ (Intercity Express Train - replacement for HSTs (manufactured by Hitachi in Kobe, Japan)) derived from sensors on the coaches. 

[JohnM]

Well you learn something new every day 

Yes, Hitachi AT300 / Class 80x IET^▸ (Intercity Express Train - replacement for HSTs (manufactured by Hitachi in Kobe, Japan)) coaches use a few complementary methods to measure carriage occupancy and seat availability:

1. Air Suspension Load Sensors (Mass/Weight Calculation)

The primary method used for overall coach occupancy on the Class 800, 801, and 802 fleets relies on the secondary air suspension levelling system.

• How it works: Pressure sensors monitor the air pressure required to maintain the carriage floor at platform height as passengers board.
• The data: By comparing the pressure differential against an empty baseline, the onboard system calculates the total live mass of the coach.
• Output: This mass is converted into an estimated passenger count (and percentage loading) and is fed to the passenger information displays (PIS^▸ (Passenger Information System)) on station platforms and mobile apps (showing green, amber, or red loading icons). This accounts for both seated and standing passengers.
  

2. Automatic Passenger Counting (APC) Door Sensors

Infrared / optical barrier sensors are mounted above or beside the passenger access doors.

• How it works: These sensors count bodies breaking the beam as people enter or exit the train at stops.
• The data: By tracking net inflow and outflow per carriage, it provides immediate, real-time footfall updates to complement the suspension weight data.

[IndustryInsider]

Well you learn something new every day 

Yes, Hitachi AT300 / Class 80x IET^▸ (Intercity Express Train - replacement for HSTs (manufactured by Hitachi in Kobe, Japan)) (Intercity Express Train) coaches use a few complementary methods to measure carriage occupancy and seat availability:

1. Air Suspension Load Sensors (Mass/Weight Calculation)

The primary method used for overall coach occupancy on the Class 800, 801, and 802 fleets relies on the secondary air suspension levelling system.

• How it works: Pressure sensors monitor the air pressure required to maintain the carriage floor at platform height as passengers board.
• The data: By comparing the pressure differential against an empty baseline, the onboard system calculates the total live mass of the coach.
• Output: This mass is converted into an estimated passenger count (and percentage loading) and is fed to the passenger information displays (PIS^▸ (Passenger Information System)) on station platforms and mobile apps (showing green, amber, or red loading icons). This accounts for both seated and standing passengers.
  

2. Automatic Passenger Counting (APC) Door Sensors

Infrared / optical barrier sensors are mounted above or beside the passenger access doors.

• How it works: These sensors count bodies breaking the beam as people enter or exit the train at stops.
• The data: By tracking net inflow and outflow per carriage, it provides immediate, real-time footfall updates to complement the suspension weight data.

I would say it's pretty accurate most of the time.  If you have a sensor failure then it can lead to erroneous readings, but that's not too common.

[Oxonhutch]

Surprisingly, if the GWR^▸ (Great Western Railway) app is to be believed, the 17:36 appears to have seats as it approaches Castle Cary.

[Image from here is not available to guests]

The preceding 17:03 to Penzance seems to only be congested in first class.

In the light of information from II^▸ (IndustryInsider - a respected member of this forum), maybe Mr Creosote was travelling first class that day. 

[TaplowGreen]

Intriguing.

So if this data is to be believed, a Friday afternoon train leaving Paddington headed for the Westcountry for one of the busiest event weekends of the year, with the previous train having been cancelled, has "many" seats available throughout with the exception of First class which is full (at those prices there must have been some very rich people on board, or perhaps the GWR^▸ (Great Western Railway) management surfing team travelling on staff passes?)

I'd say either the data has a whiff of fertiliser about it - a case of "more sensors than usual requiring repair at the same time?", or perhaps if accurate and the train was as lightly loaded as the data indicates suggests that people have given up on relying on the train and have gone by road instead - or perhaps surfed all the way? 

[TaplowGreen]

20:06 Westbury to Cheltenham Spa due 22:06

20:06 Westbury to Cheltenham Spa due 22:06 will be terminated at Gloucester.
It will no longer call at Cheltenham Spa.

It is being delayed at Westbury.
This is due to a fault on this train.

[IndustryInsider]

Intriguing.

So if this data is to be believed, a Friday afternoon train leaving Paddington headed for the Westcountry for one of the busiest event weekends of the year, with the previous train having been cancelled, has "many" seats available throughout with the exception of First class which is full (at those prices there must have been some very rich people on board, or perhaps the GWR^▸ (Great Western Railway) management surfing team travelling on staff passes?)

I'd say either the data has a whiff of fertiliser about it - a case of "more sensors than usual requiring repair at the same time?", or perhaps if accurate and the train was as lightly loaded as the data indicates suggests that people have given up on relying on the train and have gone by road instead - or perhaps surfed all the way? 

In that particular case I don't think the data can be believed.  Though it doesn't say First class is full, as the majority of First class seating is in coach K which has 'many' seats available.

From experience, recordings 'on train' are usually correct or very accurate about 95% of the time, but there's an extra level of data transfer to get loadings from the Train Management System on the train to the app, so I don't know how often that might fail.

[Chris from Nailsea]

Thanks, as ever, for your useful 'insider' information here, IndustryInsider.

CfN^▸ (Chris from Nailsea, an administrator on this forum). 

[grahame]

17:50 Gloucester to Salisbury due 20:06

17:50 Gloucester to Salisbury due 20:06 is being delayed at Gloucester.
This is due to train crew being delayed.

[TaplowGreen]

20:06 Westbury to Cheltenham Spa due 22:06

20:06 Westbury to Cheltenham Spa due 22:06 will be starting late from Westbury.
This is due to the train departing late to maintain customer connections.

[grahame]

20:06 Westbury to Cheltenham Spa due 22:06

20:06 Westbury to Cheltenham Spa due 22:06 will be starting late from Westbury.
This is due to the train departing late to maintain customer connections.

Our old friend the 18:36 Paddington to Plymouth, due Westbury at 19:58, didn't arrive there until 20:09.  Sensible hold on the 20:06 which left 6 minutes late.