Reports coming in about an incident involving an SWR and a GWR service in Fisherton Tunnel near Salisbury.  GWR reporting disruption until "the end of the day" due to the emergency services dealing with an incident.

 
Now saying disruption until the end of tomorrow (Monday).

 
From Wiltshire Today

 
From Wiltshire Police on Twitter

 
And now from GWR

 
GWR just tweeted that the two trains involved were the 1708 Portsmouth Harbour - Bristol & the 1720 SWR service WAT to Honiton.

One news source I saw said that the GWR service has partially derailed with the driver trapped

 
From the BBC


Story continues in link above.

 
“Close to Grately?”  Some sort of error with initial reports I can only assume. The Twitter link no longer works anyway.

 
There are some  photos posted around.i think we are very lucky we are not talking about fatalities in this incident especially when you see the stage of one of the SWR units. 158763 going to be put of action for a while too.

I'm not going to post the photos as I don't think.its respectful

 
And still no "Passengers between Southampton/Portsmouth and Cardiff/Bristol with tickets routed via Salisbury should travel via Reading".

 

I don't think it's particularly relevant to talk about how long one of the units is going to be out of action in an incident that had the potential to be life changing for those involved (although at this moment it thankfully appears as though passenger injuries may be minor  - not yet clear on the driver who needed to be cut free, and my thoughts are with him at this time).   

 
From the BBC


I understand that the driver has now been cut free (was a while ago but it takes time for news to flow)

 
What I find shocking is not the FIRST collision, between the first train and and an "obstruction". That sort of accident can never be entirely prevented.

A second train then collided with the first one. I would expect that the signaling system  would prevent that. The signaling system would "know" the location of the first train and keep the following train a safe distance away. I appreciate that the initial collision reportedly put the signals "out of order" But signals should fail safe. If the relevant signal was out, that should still be treated as being a signal at danger.

Or was the second train moving on the other line ? That could explain the second collision if the first train was foul of the line.

It has been suggested in some reports that a landslip or similar caused the obstruction.

 
The RAIB are on site and I suspect the signalling system will be one of the things they will be investigating as a priority.

 
Having seen various pictures from the scene it is remarkable that there weren't more serious injuries or deaths.

The incident has rather affected my plans for today. I had planned to visit the Isle of Wight to ride the 'new'  Class 484 stock and see the station improvements.

Just debating with myself whether Templecombe - Exeter - Reading - Basingstoke - Portsmouth is a reasonable route as allowed by SWR's ticket easements today. There's no news yet on rail replacement bus services in the Salisbury area. Yeovil - Dorchester - Southampton - Portsmouth already has rail replacement for engineering works in the Brockenhurst area. Also my last experience of a rail replacement bus was Finn being denied travel.

EDIT. SWR Twitter just confirmed I can travel via Exeter and Reading. Long way round but avoids buses.

 

Not just the signalling, the GWR train driver eye witness evidence will be key also the train black box and Forward Facing CCTV (if fitted), the physical condition of the track, tunnel etc.

 
Like Broadgage, I am really concerned at the second collision ... being careful not to speculate too far.

Good to see everyone is out and alive - hope the injuries to the driver (in spite of him (?) being trapped) are not life changing - I am keeping an ear open for a medical update and will post when I see one.

The Waterloo to [Honiton/Exeter] hourly service does - of course follow the Portsmouth to Cardiff service through, as per yesterday evening, making an excellent connection from Southampton to the west, and utterly failing to connect for Waterloo and Clapham Junction to Trowbridge / Bath / Keynsham / Bristol passengers.

 

It does sound as though injuries were not serious, which is wonderful news if accurate. I assume most, if not all, passengers would have been seated, which would have helped. Regarding the driver, one can be trapped following an accident without being hurt, and I hope this was the case here.

 
The two services on Real Time Trains. The second train not detected at Tunnel Junction. Both running late, the SWR service making up time, and both arrived at the junction with not many minutes between them. Thoughts with everyone affected.

Mark

https://www.realtimetrains.co.uk/service/gb-nr:G57571/2021-10-31/detailed

https://www.realtimetrains.co.uk/service/gb-nr:L52648/2021-10-31/detailed

 
Real Time Trains data only there for a week "after".  Thanks for the links, Mark - here are the pertinent bits:

 

I have seem some reports that there were people standing as they prepared to get off at Salisbury.

 
Looks like 1F27 passed though by the skin of its teeth!



Other notes (from BBC Wiltshire) ... 3 remain in hospital. Driver serious but comfortable. "About 100 people on the train(s)". Can't trust everything you hear - comment from Claire Mann referred to as "Managing director of Great Western Railway" ...

 
Guard on my train from Templecombe to Exeter has told me that no units have been allowed out of Salisbury depot this morning. That has meant only a shuttle service can be run between Gillingham and Exeter using units that outstation overnight at Yeovil and Exeter.

One of those units is parked up at Yeovil Junction, low on fuel, having not been able to make it back to Salisbury last night. It may have to be fuelled by road tanker if it can't be taken back to Salisbury in the next day or two.

Question. Can SWR fuel their 159s at Exeter if need be?

 
From overnight:

 
There are now enough daylight pictures (e.g. on the BBC) to confirm some of what was "known" but confusing last night.

The eastbound GWR train (1F27 ) passed Salisbury Junction first (timed 18:41 - 5 minutes early). This is a sharp right turn, limited to 20 mph. For reasons still unclear, this train was held at Dean (some distance before the station) for over two hours before departing 151 late and being cancelled at Fratton "Reason TY [Mishap-Train Operating Company cause] ".

The westbound GWR train, subject to the same 20 mph speed limit round the sharp left turn, was timed at Salisbury Junction at 18:42, 19 late. It stopped with the last vehicle outside the tunnel, and still on the points. Its rear end was thus well to the left of the Down Main Line that 1L53 was arriving on. This last vehicle was reported to have derailed, but I think that might be a mistake. It was struck soon after so ended up derailed anyway, and it does seem unlikely that an obstruction would derail only the last vehicle - and it never entered the tunnel. We'll have to wait and see on that one.

The SWR train (1L53) had earlier been cancelled before departing "Reason JP [Failure to maintain vegetation within network boundaries in accordance with prevailing Network Rail standards]". It was reinstated and was last timed departing Andover at 18:30, 6 late. It should have passed Salisbury Tunnel Junction at 18:39, so with the delay would be expected at 18:45 (any time in RTT is interpolated). Its straighter run over the points allows a 50 mph speed limit, but it should have been slowed, delayed, and then stopped by the signals. The biggest question at the moment is why that did not happen.

We know 1L53 bit not stop, and struck the rear vehicle of 1F30 a glancing blow. It is not possible to tell from the resting positions where that stopped carriage was when struck, but in its current position the two trains would not have passed to strike corner to side. The impact would have been corner to corner, and much more violent as a result. Either way, the leading end of that GWR vehicle was did not pass the tunnel portal; it's not clear from the picture whether it lodged against it and could go no further. 1L53 was thrown off its track, and half overturned, and ran into the tunnel in that canted position along the Up Main and (at least in places) became wedged between 1F30 and the tunnel wall.


Two things went well in the accident, accounting the the relatively light injuries, and to only a small number of passengers and staff. One was that the body shells largely kept their shape and bounced off each other. The other was that no fire broke out. It could have been very much worse.

 
If you lighten the image then there seems to be something between the two GWR carriages just at the tunnel portal.

 
BBC lunchtime news reports 5 people still in hospital, all described as stable.

 

Now it is being reported that the SWR driver's injuries are more serious, and sadly will have long-term consequences.

 
I think in Stuving’s excellent summary the eastbound GWR (1F27) would have been stopped by an area emergency call.  I’m therefore assuming that if they were still in Salisbury’s control area then it would not necessarily be a priority to get them going again. 

But what happens to the signallers who are on shift in such circumstances, do they have to get relieved?

Paul

 

Looks like the impact zone to me, which would place both trains further back from the tunnel mouth at the point of collision.

 

Alas so, using that awful term "life changing".

 

As soon as practically possible, yes.  Especially if there’s any reason to believe their actions may have caused or contributed to an incident.

 

Yes, that is all in the Salisbury signalling area. That explanation would explain the initial stop, and I wonder if there was a concern even that early to get information from the driver about anything on the track, or any other source of evidence such the FFTV.

 
Apparently the Telegraph (paywalled) are now reporting that there was only one incident:

BTP on Twitter:
https://twitter.com/BTP/status/1455173517847171074?ref_src=twsrc%5Egoogle%7Ctwcamp%5Eserp%7Ctwgr%5Etweet


Also supported by a recent BTP statement:
https://www.btp.police.uk/news/btp/news/england/derailment-at-fisherton-tunnel-salisbury---statement/

 
Eyewitness account from a passenger travelling from Brighton to Southampton and then onto the GWR service where he was to alight at Salisbury - and in the event ended up being walked through the tunnel and along the tracks into Salisbury Station which is quite a trek in itself.

It may be significant that he doesn't mention two events.

https://www.salisburyjournal.co.uk/news/19685161.salisbury-train-crash-passengers-experience/

 
Thoughts with all rail staff, emergency staff and passengers who were victims of or witness to the accident, especially the driver, whose injuries seem to be more serious than first reported. Thank goodness no lives were lost.

 
If I have understood correctly, the two trains involved were travelling in opposite directions.
In that case, I retract my earlier remarks that imply that the signalling system should have prevented the SECOND collision.

If two trains are proceeding in the SAME direction, on the same line, then even if the first train suffers an accident, the second train should be held a safe distance away by the signalling system.

If however two trains are proceeding in OPPOSITE directions, and the first one is derailed, then it can be hard to prevent a second accident. If the first train, or any substantial parts thereof is foul of the OTHER line, then no automatic system can reliably prevent the second train colliding with the the first one.

Under such circumstances, reliance must be placed on human actions such as;
Press the emergency button on the cab radio, so doing sends a "stop danger" message to other trains in the area.
Apply track circuit operating clips to other line, this puts signals back to danger.
A member of train crew to walk back along the track with a red flag or lamp, and apply detonators.
Make an emergency phone call, preferably from a signal post telephone. Cellphone as a second choice.

Such actions might not be in time to avoid a second accident.

IMPORTANT EDIT. Both trains were travelling in the sane direction. See following post for details.

 

The two trains were both travelling toward Salisbury and collided at Salisbury Tunnel Junction. The GWR service was travelling from Portsmouth to Cardiff and the SWR service from Waterloo to Honiton.

 
There are now more recent reports this afternoon, eg from BTP that I quoted, that there was no second event at all…

Paul

 
One conclusion from the published pictures is that both trains were moving when they came together and collided. Note that the trailing cab of the GWR train shows no sign of being hit by the leading cab of the SWR train.

 
That BTP press release of 13:40 today starts off with:

That would fit in with the idea that both trains were moving, though that in itself raises problems over the recorded timings. But I'm still not taking it as a definitive statement, given some bits that are are obviously wrong.

 
There are informed suggestions that the GWR train split in the tunnel - so the picture showing two trains from the front roughly side by side is of the second of two 2-car GWR 158s. When that happened isn't clear.

 
‘Travelling in the same direction’ simply means they were going the same way (towards Salisbury station), not that they were actually both moving.

We’ll have to await the initial RAIB report to know for sure. I suspect also that the BTP will be ‘asked’ to shut up about what they think happened & await the report too. They were wrong with an initial report about the derailment prior to the crash….

 
From GWR, SWR and Network Rail, by email, headed "Incident at Fisherton Tunnel"

 

They have been stated as "life changing" but just stated as "life threatening" in a BBC report I have just heard. Let us just hope that's a slip of the tongue; a difficult day to get wording accurate in a sea of developing and amending news.

 
POST EDIT, 02-Nov-2021: In view of the RAIB's information today that they are considering wheelslide as to why the SWR train collided with the GWR train I have removed what remained of the post I had put here.

Dave

 

Can we please refrain from speculating! Let the RAIB investigate and publish their conclusions in due course.

 

Let’s say this is correct, and the front portion continued past the signal outside the tunnel. Would that then release the previous signal, allowing 1L53 past the junction signal at amber? Just one hypothetical based in what’s been reported here so far. Yes, it raises other questions of course, but possible. There are zero photos of the front unit of 1L30, which raises my suspicions that more has still to be discovered. A lot more possibly

 
Indeed but I resisted the idea of expanding further and so haven't said precisely what I am thinking. A colleague of mine at work is a former NR employee though and he and I were talking it over earlier today.

 

You said enough - and indeed stated that ' I won't post my conclusions here (although some may guess them)'

A little knowledge is a dangerous thing!

 
In my view, speculation is fine provided that it is firstly reasonably well informed, and secondly is based on previously known facts about infrastructure, rules and procedures, and rolling stock, and/or based on reliable reports already made public by reputable media, such as published photographs of the aftermath, or published statements and press releases.

Speculation blaming an individual is in poor taste and might in extreme cases even be illegal.

But more general speculation is arguably at least partly what forums are for. The post by  DaveHarries is in my view informed speculation, incorporating factual data about infrastructure in the area, and is in my view both interesting and informative.

EDIT TO ADD that the post to which I refer is now hidden from public view. In my view it should be re-instated, but ultimately the view of the moderating team must be respected in such matters.

 
It may be worth further discussion….I’m also happy with discussion on infrastructure while not discussing persons and/or staff (so no personal stuff either persons or positions with the rail industry) which both my previous post & Dave’s was

 
There will be no public speculation from me.  I await the RAIB’s findings with interest.

 

It is a very difficult call indeed for me, for the reasons you have stated.   I have stated in my "snip" comment that I am seeking the views of the wider moderator team. Its removal is precautionary at present; it can easily enough be put back, if appropriate, but if I left out in place while we decide, those extreme cases of legality you refer to could cause permanent damage.

 
As will I. My earlier post is hidden but I will keep my conclusions to myself. Apologies if I said too much: I will, of course, read the RAIB report also with interest and will be happy if I am proved wrong on what I am thinking.

Dave

 
Thankyou Dave.

 

I fully agree with your post.

The likes of the RAIB carry out their own thorough investigations, highly professional people. Not the kind who give up, retire to the pub trawling through public forums to copy and paste posts to compile a report.
Point the finger of blame at someone or naming individuals crosses the line but speculation based on some knowledge is in my opinion, perfectly acceptable, a lot of us have discussed and speculated all day in messrooms today.

 
The problem with speculating publicly is that there are plausible explanations that point to human error.

 
No worries although I remain hopeful that the hiding of my post is only temporary. If I had gone the whole hog and given my conclusions in the post (which I won't do on this forum or anywhere else public) then I would fully understand the hiding and / or removal of the post permanently. I realise that it is not my say-so though.

Dave

 

True, as with any incident there’s always going to be the possibility that human error plays a part to  a varying degree, but where human error occurs a safety system like TPWS or ATP should prevent incidents of this nature which is why this one has caused a good deal of debate.
I think we can all appreciate that no one sets out to deliberately cause an incident.

 

Indeed, and we live in a society which very rarely says "it was an accident - just one of those things that happens from time to time".  We ask which individuals, and which systems, could have done things or been different and we then ask "could they have know better" or "should the systems have been better". And as such, we tend to point to human error or misjudgement; doing so without full information tends, perhaps, to create something of a kangaroo court.

Looking back even at old reports I can turn the pages of virtually any of them and read "x" should have done "y". Once the verdict is in, I can quote that. Prior to than verdict / analysis, I really shouldn't even draw a parallel and indeed I'm going to stress here that the reports in this photo are merely the more southern ones from a pile I have.




Indeed - but there are some accidents which, upon proper investigation, have turned out to have at least partial cause based on wilfully not following rules which are laid down for safety reasons.

 

One which immediately springs to mind is the one bottom right in the picture. Very much wilful rule breaking in that incident.

 

I think wilful rule breaking is extremely rare these days compared to back in the 70’s. Back then the work was tough but it was a job for life, you could get away with incidents.  Today, there’s been a reversal. Actual driving is far easier but it’s now significantly easier to lose your job. For a driver, I’d say it’s no longer a job for life. Wilful rule breaking today will get you caught, you’ve got your CCTV cameras, OTDR etc.

 
Is Anyone able to explain why SWR trains are unable to get on and off the depot at Salisbury?

I don't know the area very well but looking at Google satellite view and open train times its not obvious its a long way from the tunnel to the station area.

 

looks like all the area controlled by Salisbury panel signalbox is out of use, pending investigation being complete?

 

Lots of thought given to this.  I have outlined my concern about publishing speculation on the forum while we await (and it is frustrating!) the facts and findings and it's been discussed amongst the moderator team. Not only are we looking at legal issues, but also at hurt potentially caused to those involved and the possibility of rumours and shares of what turns out to be wrong ... indeed we have witnessed, in the last 24 hours, initial reports of the incident starting with a derailment caused by an obstruction being stated "no such evidence" by the BTP.

All of our members who have posted a few times have access to our "Frequent Posters" board.  It allows established members somewhat more freedom, and we require that members with access to that board do not share its material any wider.  The issues that caused us to snip the public post above would be mitigated were such a post to be made out of the public (guest) view, and mitigated further by "Frequent Poster" guidance.  That is far from saying "anything goes" - it does not - but I would remind members of the availability of the member's board and invite appropriate (re)posts of some material there.  I hope this provides you with a way forward.

 

I can’t see that being the case, as trains from Cardiff to Salisbury and vv were running all day Monday, and Salisbury panel would have to be controlling those train movements between Beechgrove (just outside Warminster) and Salisbury.

 
Trains running today Exeter to Salisbury.

 
BBC South Today at lunchtime was quite categorical - RAIB are investigating why the SWR train passed a red signal, and looking at the condition of the rails.

The report did try to distinguish between BTP, controlling the site, and RIAIB doing the investigation work. I have an idea that it suits RAIB fine to work behind the police, and in particular they are happy never to meet a journalist. They accept that the police will want to know if anything points towards a criminal investigation, and intervene if they are concerned about conserving evidence according to the rules.

More upbeatly, the SWR driver is now reported as less badly injured than was feared.

 
RAIB announced preliminary result
Seen here on Twitter
https://twitter.com/philatrail/status/1455568926469808131?s=21
Website source
https://dft-newsroom.prgloo.com/news/raib-statement-salisbury-rail-accident
Edited to give DfT website source

 
In thread post as an alternative to earlier attachment.

RAIB Statement: Salisbury rail accident​
2nd November 2021

 
Leaves on the line then?

 

Or perhaps an insufficient number of RHTT workings over that stretch of line in the days leading up to the collision?

 
As I suspected. Unfortunately this highlights an issue with the TPWS system.

 
Yes, looks like the root cause is wheel slide then.  Probably what most of us within the industry suspected.

Further investigations will no doubt check whether the railhead treatment plan for the location was appropriate and followed properly, and that the driver was driving the train up to the point of impact as per SWR instructions (and whether those instructions are appropriate).

 
Really got to feel for the driver. Must have been terrifying for him.

 

I hope they look at the part TPWS played in the incident. Once the train passes the red the driver loses all control and is effectively a passenger. I’d be interested to know if the emergency brake application made the slide worse.

 
From the photos/pics I can find, the signal protecting the junction is well back before the tunnel & seems to be the other side of the A30(?) overbridge. A hecknof a long way to slide? Am I right? I understand that the speed limit from the BSK direction is 50mph, and the damage I estimate must have been going almost as quickly on impact to carry the train that far into the tunnel

 
Falling gradient too, I believe.

 
I wouldn’t describe that as a heck of a long slide at all, and I certainly wouldn’t want to make even the vaguest of guesses as to what speed the impact was based on a couple of photos!

 

I managed just under 2 miles several years ago with all wheels locked. Luckily plain track with no restrictive aspects, applied the brakes at about 60mph, a mile or so later, I hit the platform ramp doing 55mph.

What distance from a signal protecting a junction to the junction is safe, that would prevent incidents like this?

 
You are indeed. A look at the street-level view from the A30 overbridge on Google Earth shows the protecting signal - SY31 - in one direction while rotating the image 180 degrees allows a small view of the location where the derailment and collision occurred.

Dave

 
Thinking of Google Streetview, the condition of the cutting sides over the last ten years ago contrasts sharply with their appearance in, ok, 1970, which I know is a different era, but they were clear of scrub and trees (photo from Flickr). The allotments to the right are still there, but even they are now losing out as they too are shaded by the full grown trees at the top of the cutting.

The google streetview is from June 2015 and provides the human interest of a couple of people in orange hi-viz gazing at the tunnel approach.

Given the two converging lines, junction and tunnel, there's something to be said for returning the cutting sides to grass. (And there's a lot to be said for properly resourcing the Cardiff - Portsmouth, the Waterloo - Exeter and the Waterloo Bristol traffic flows so that they're not miserably capacity-constrained for decades on end and can contribute properly to underwriting the necessary work to the infrastructure. As Graham's said, simply ensuring useable connections between Waterloo-Salisbury and Salisbury-Bristol might quickly find passenger numbers rising to 240,000 a year. The reason it's not done is presumably that it would overwhelm GWRs provision between Salisbury and Bristol...)

https://goo.gl/maps/25yDmdeDFB5fcz8s5

https://www.flickr.com/photos/96859208@N07/10873405094

Mark

_{Edit: replaced first link with shortened version - Red Squirrel}

 

The scheduled RHTT run for earlier in the day on Sunday was cancelled 'at the request of the train operator' and doesn't appear to have run in the days beforehand either. It would have passed through Tunnel Junction four times.

 

Ouch!

A member of the in-law contingent managed most of a mile at the controls of a Blackpool tram some years back, including the on-road bit by the Metropole hotel, coming to rest by the Tower. All he could do was sound the horn. He didn't hit anything or anybody. It was ice, not leaves, but shows the effect of lack of adhesion. The big question was why it had never happened before in the previous century, given that the conditions on the day were far from unheard of.

 

Blimey. I bet m'learned friends will already be mobilising given the way that vegetation has been allowed to proliferate with consequent leaf fall on the lines - assuming it's Network Rail land of course.

Similar scenes of hugely overgrown land near the railway all over the country of course. I wonder if this will catalyse NR to do something about it?

 

For posterity, here are extracts from that URL

 
Any idea why the train operator would ask for it to be cancelled?

 

No driver, no operators, loco failure, set failure, redeployed elsewhere…..

 
This cab-view video shows the state of the vegetation in the cutting west of the tunnel in 2015.

In this video, signal SY31 is passed at around 38:33 (having become visible at around 38:21), and the junction is passed at around 38:45.

https://www.youtube.com/watch?v=FLV_D2RHeMc

 

apologies - typo -  I meant *east* of the tunnel, not west.

Also very best wishes to the driver https://www.bbc.co.uk/news/uk-england-wiltshire-59143021

 
Driver is 75…..is s/he the oldest on the railway?

Various media also reporting a broken ankle

 
There’s an updated RAIB “news story” this morning that explains current understanding:


https://www.gov.uk/government/news/collision-between-passenger-trains-at-salisbury-tunnel-junction

 
No doubt the full report will focus intently on why the RHTT didn't run as scheduled. There may be some difficult questions for Network Rail to answer.

 
The RIAB statement has this to say about the approach by 1L53 from further back:


The caution signal, SY29, is passed on that cab video at 37:55, followed by the 50 mph sign just by Laverstock North Junction at 38:08. I presume slowing for the speed limit reduction (from 90 mph) would need to start well before SY29, and once its aspect was seen the speed limit made no difference.

I still can't quite get the timings to work and place the two trains on the junction at once. It's only a couple of minutes, so it may be down to the nature of the timings recorded, or indeed to my sums. The distance from the call at Andover to Salisbury Tunnel Junction is 16m 13ch, and at a steady 90 mph with no call would take 10.9s. Add acceleration time and it's barely possible to get there in 12s (with rounding of times) and the path allowed 15s. Recorded timings at Andover and the junction are 18:30 and (for 1F30) 18:42, so by my reckoning there can't have been much deceleration at all.

Thus I was not surprised by the RAIB's comment about slide starting from the initial service braking application. As to why ... well, that stretch of track is not overhung by big trees like the junction itself. So I shall be interested to see not just the actual timings, but the full explanation of the lack of adhesion. If there is one, of course; previous investigations have often failed to come up with clear evidence from the railhead itself to fully explain sliding.

 

I went to school with someone with the same name as the driver: not a particularly common surname, right age, right area.................I wonder ??

 
Whilst leaves on the rail are the most probable reason for the reported wheel slide, I suppose that oil or grease on the rails is an outside possibility. That could result from a defect on another train leaking oil. Rare, but just about possible.

ISTR a buffer stop collision that occurred some years ago despite the driver braking correctly. Oil or grease contamination was considered as a possible cause.

 
I think your question is overtaken by the latest RAIB explanation, the driver was already emergency braking, so AIUI in that case TPWS just duplicates the demand, but there’s no additional effect?

Paul

 
On my train from Salisbury to Templecombe earlier today it was noticeable how much more cautiously we approached each station stop. I presume drivers have been asked to drive even more defensively than they usually do.

 

It would seem that the combination of Mk1 human eyeball and Mk1 human brain beat the computers to it.