[Update 2: A reader sent in the scanner traffic reporting the incident. Listen below.]
[Update: Sound Transit worked through the night to fix track and signal damage, and service returned to normal Wednesday morning.]
Around 8:30pm Tuesday a person driving a white Chevy Tahoe SUV collided at high speed with Link at Holgate Street in Sodo. Reports from the scene indicate the driver drove around the crossing gates. The vehicle caused significant damage to LRV #136, collapsing doors and exposing insulation. A heavy police and fire response indicated that there were passenger injuries, but fortunately the vehicle seems to have collided with the luggage/bicycle storage area, minimizing injuries. As far as we know this is the first time a vehicle has penetrated the passenger areas of a Link train in regular service.
Photos from the scene are harrowing, but details are sparse at this hour. We hope to have more info in the morning.
Until further notice, likely through the end of service Tuesday, service is suspended between Beacon Hill and Stadium, with a bus bridge in place.
https://twitter.com/staceyhanson/status/834272556852981760
Take away this person’s license. Permanently.
Throw that A-hole in jail and never let him drive again.
On the operational side, this is where interchangeable LRVs come in handy. Hopefully the other LRV car in the train (that wasn’t hit) is still operable.
Use the 100-footers on the lines with at-grade segments, and the driverless open gangway supertrams on the totally grade-separated lines.
Lets stop apologizing for LINK’s problems here. It’s trying to do too many things with no clear direction. Is it a light rail line serving the city? or is it an interurban line, that needs high capacity and direct service to serve the suburbs? What is it? Once that question is answered, lets start solving the capacity issue. Rainier valley bypass, Check. expanding platforms for 6 car trains, Check. That’s how you solve the real issues, not by Band-Aid approaches with new, expensive, custom rolling stock.
>> [does it need] high capacity and direct service to serve the suburbs?
What a silly notion. If we ever have a capacity problem, it won’t be because it serves the suburbs (directly or indirectly) it will be because, like every other transit system ever built, too many people in the city are using it. Spending billions for a light rail line on a commuter rail/express bus corridor is crazy enough — building two of them would be nuts.
We just voted on Link’s purpose three months ago. It’s a hybrid system that serves the suburbs, most (but not all) of the city’s urban villages, and other areas that happen to be 1-2 miles along the way. The Ballard and West Seattle lines were specific commitments to the city beyond the central spine. I don’t see what this collision has to do with capacity issues. Any current capacity problems are simply that ST didn’t order enough trains for 2016. In the future if any part of the system reaches capacity it will be between downtown and Northgate. Siome people are concerned about that but it depends on which predictions you believe. If it becomes an issue, ST can dust off the project to retrofit the DSTT for sub-3 minute headways. Or we could get on with that Aurora line. The Georgetown bypass for Rainier Valley doesn’t have enough support: it was in ST’s long-range plan but deleted in 2014. The supposed beneficiaries of it, South King and Pierce, didn’t even ask for it in ST3, nor raised any concern about south Link’s travel time.
Those denying the existence of a vision for Sound Transit are really just disagreeing with Sound Transit’s vision. It’s not as if we voted for Electors to the Sound Transit Electoral College. We voted for a package. It passed. Let’s move on for Creator’s sake.
Bill the driver for the repairs to the train car, plus all the costs of the bus bridge, plus the time of every passenger delayed at $15/hour.
Oh, and take away his license, too.
It’s a multi-million dollar moment of impatience on their part. ST had better be made whole by their insurance, assuming they have it.
Yes, very much so. Buy ST a new Siemens S70 train. This one’s total’d and belongs in a children’s museum.
This is anger-inducing on so many levels, I have to remember it is times like THESE we have a good comment policy ;-).
If it really did cause multiple millions of damage, the driver’s insurance policy won’t pay anything near that. Auto policies have limits well below that.
It would have to go to court to determine that. In the case gross negligence is determined sometimes they wind up having to pay anyway.
The light rail car shouldn’t be that difficult to repair. It’s sheet metal and a new door in that area. The big time consuming thing will be putting the new door on and trying to get it aligned with the door closing mechanism.
What could make things expensive are the injuries to the passengers.
When a BMW came crashing down on top of the Cedar Hills Transit Center, it only cost TriMet $58,000 to replace all the overhead lines. The driver was the only one that was around to get injured in that case though.
Insurance guy here. There’s a very low chance this driver has sufficient insurance to cover ST’s costs (could easily be $1M+ considering labor costs and damage to equipment).
Property damage liability limits on auto policies can be as low as $10k in WA. Rarely are limits above $100k.
Beyond the Auto limit a separate Umbrella policy would be required to cover additional damages. That’s not very likely.
Bottom line, ST is likely on the hook for a lot of these costs. I don’t know how ST structures its own insurance but high-deductibles are pretty common for governments and large companies.
Of course, ST has insurance. Like most big agencies they’re probably self-insured.
The driver of the Tahoe was likely “in a hurry” and thought that they would save five seconds by going around the gates. The cause of the collision: My guess impatience and speed.
“The check for $58,181.03 from insurance company USAA covered repair work to the Sunset Transit System and the cost of providing bus service for MAX riders, TriMet said.”
For the record, at-fault drivers can be held liable for added operational costs. Thanks for pointing this out, Glenn!
Is it just me or does this seem like a bit more than impatience? Driver had to blow through a gate that was down and lit. The damage looks like SOV didn’t bother to apply brakes and if its a T-bone the side of the train was in the front windshield.
Wouldn’t surprise me if this was a wacky protest to his car tab fee.
It’s extremely fortunate no one was hurt or worse. I wanna kick that driver for wrecking #136.
That should read seriously hurt.
Grade separation.
Exactly.
Does anyone have any back-of-the-envelope estimates for what it would cost to separate Link from car traffic in Rainier Valley? I think I’ve seen people raise the possibility of converting the intersections to underpasses for cars, which would presumably be considerably less expensive than rebuilding it as entirely elevated, since you wouldn’t have to build new stations.
I’d be surprised if any option pencils out as low enough cost to be a high priority, but I would be interested in seeing any reasonable numbers.
More than it cost to build the first time, because there’s the added complexity of building around an operating line. Billions.
ST4 folks should be Seattle centric to fix these kinda messes. Period. Taxes raised in Seattle only to fix Seattle only.
It really depends on what is under there that is in the way. A major water or sewer line is a huge obstacle to deal with. If something like that isn’t under there then it might not be too bad.
In a rural area it is easy. You wind up digging a pit under the line and you get this:
https://www.google.com/maps/@46.585065,-123.2860347,3a,75y,239.78h,91.62t/data=!3m6!1e1!3m4!1sJIB-Xpu3cYXwY-3aPacykw!2e0!7i13312!8i6656!6m1!1e1?hl=en
In places such as Kent, you get this:
https://www.google.com/maps/@47.4412109,-122.2395369,3a,60y,262.16h,84.94t/data=!3m6!1e1!3m4!1s-fQLGwCeiqRXAnltyzXO8g!2e0!7i13312!8i6656?hl=en
The question is how to handle the intersecting traffic on ML King though.
My completely amateur opinion is it would be better to elevate the line in RV over the existing ROW. You should be able to (mostly) build the elevated line while not disrupting existing service. A full elevation not only removes the grade-separated street crossing (boosting speed, frequency, and safety), but it also reconnects the neighborhoods where Link closed off dozens of minor street crossings.
But if that’s too expense or has other issues, digging a handful of trenches is a good solution.
@Dan I see your point about building while operating the existing line adding to the cost.
@Joe While separating Link from cross-traffic would have the greatest benefit for Ranier Valley residents, it would also benefit anyone going to/from any Link station south of Beacon Hill. That includes people going to the Airport/Angle Lake/TIB stations now and Tacoma/etc in the future.
@Glenn Do you know about how much it costs to build something like the Kent underpass you linked to?
@AJ While rebuilding the Rainier Valley segment as elevated would be great, constructin 4 new elevated Link stations would also add at least a couple hundred million dollars to the price tag.
Rather than spend billions to elevate the tracks through SODO and RV, a better solution might be to spend an order of magnitude less to build an alternate line along the mainline RR tracks next to I-5 to the S. Boeing Access Road and join the existing line. That way there could be express trains to the airport, and an alternate route trains could use when there are disruptions. Having choices is good, right?
FWIW, I was just in Texas for a weekend and they have those sort of trench underpasses all over Austin
@Philip – for sure it would be much more expensive, easily $1B range all in, but I’d argue it’s a worthwhile investment for SE Seattle.
Building underpasses for the cross streets would never fly in a built-up environment. The new bridge for the rail would be on the order of 3 feet deep (bottom of beam to existing grade. Then the roadway has to go low enough to provide 14′ clear for trucks. This requires 340 feet of ramp at 5% grade, plus `20-foot transition zones at the top and bottom of the slope on each side. Add in the width of the double track and you need approx 800 feet from the beginning of the cut to the end.
You also need to lower MLK Way (again at `400 feet on either side) to make connections with this new, lower road or it would become a limited access highway. So now how does one access the shops and housing in the blocks that front these streets? You’ve effectively created cul-de-sac neighborhoods that can only be entered from points away from the major intersections.
This also is true of the rail stations themselves. They are all placed near a major intersection. If said intersection is lowered 17′ how do you get people away from the stations?
“an alternate line along the mainline RR tracks next to I-5 to the S. Boeing Access Road”
That was in ST’s long-term plan as a possible future project but it was deleted due to lack of interest. The bypass would speed up trips to the far south end but it wouldn’t obviate need to serve Rainier Valley and Beacon Hill, which have four urban villages. Even if Rainier Valley became a shuttle line, the same problems with a surface alignment or changing it to elevated or a trench would remain.
Good points, Pete.
I didn’t realize it was in the LTP at one point. It’ll be interesting if South King agitates to get it back on the LTP once they start riding Link in significant volumes. Tacoma would probably be more interested in improvements to make Sounder faster.
Mike is right – a bypass fixes the speed issue for through-riders, but the actual local issues with safety and connectivity in RV remain.
Four quadrant GATES. And close Holgate; it’s a very minor cross-street there.
Holgate is one of the few East-West routes out of Beacon Hill – it’s ringed by both I-5 to the West and I-90 to the North. Holgate is a major thoroughfare for residents, being one of the few I-5 flyovers. Close Holgate and you’re left with Spokane, then Colson .
Keeping the flyover and capping it at Sixth doesn’t make sense either.
-Beacon Hill resident who remembers the last time Holgate was closed for a significant period of time
Holgate is definitely a major street for this part of Seattle as Nick H wrote. I’d like to see the Lander Street overpass project deal with not only the BNSF tracks but also the Link tracks. BNSF is a few blocks west of Link at Lander but avoiding trains that block the road every few minutes seems like a worthy project (plus buses in the busway). There was some talk about a Holgate overpass in years past: http://www.duwamishtma.org/uncategorized/lander-street-overpass-a-nogo/
How many people travel between Beacon Hill and Fourth or First Avenues between Lander and Royal Brougham? That is the entire population that would be hurt by closing Holgate at the Link tracks. Folks can still use Holgate to get off the hill to downtown Seattle or West Seattle simply by turning one way or the other at Sixth South.
Grade separation does not cure stupid.
It makes entire classes of collisions infeasable. A car can’t climb up to an elevated track or go down into a tunnel. Pedestrians would have to climb up to the track, and clearly trespass to do so. It’s not a shortcut or something people can do casually just walking across.
The human and financial cost of collisions should be included in the cost estimates of surface alignments. Then they wouldn’t look so artificially cheap compared to grade separation. On top of that, grade-separated trains can run at full speed while surface trains are limited to the speed limit of the adjacent streets. That translates into fewer runs per hour, which is inefficient and sub-optimal.
I can’t help but notice there’s been are exactly zero collisions between Link and cars downtown, and I don’t think that’s because the drivers are smarter.
Because one person was so impatient to get to his destination we have several injuries and one busted up LRV. Throw him in jail!!
Forget grade separation. How about a double crossing gate that blocks the oncoming lane?
+1
In the Rainier Valley too please!
First, my concern is for the passengers & crew affected by this senseless attack. I will spare this blog what I really think (try something very bloodthirsty after listening to three hours of Todd E Hermite hate and spile about the GREATEST transit agency on the face of Earth).
Second, I hope the train can be declared total’d and Siemens can speed up getting a replacement train out here. Hopefully the hulk can be restored for say the Pacific Science Center or something like that.
The damage appears to be relatively minor. less than 30 working days in the bodyshop and it will be good as new. Once all the parts are in, 2-3 days to remove the damaged portions, 2 days to repair structural, 1 day for carbody sheetmetal, 2 days for replacement doors, 4-5 days to for skim coat, primer and paint and 2-4 days for interior repair to walls, partitions, and floor. So roughly with two persons working, 18 days of actual work, labor charge if billed at around 90/hr which dosent seem too far off from what I’ve seen for that field, $25,000 for labor, plus parts. That’s assuming the frame didn’t get bent or anything else weird happen. They do (did?) have a spare end section floating around, I think it got used on 102 at one point in time so they could strip the electrical out of it and rebuild It that way if they had to…
Well Relatively minor as far as major damage goes. I have seen buses get new front ends in in around a month’s time for comparison purposes. That includes removing the collapsed front end, rebuilding the framework, installing the new fiberglass front, and replacement of wiring and electrical, a trip through the paint booth and back on the road.
This certainly sounds like reckless driving (yes, that’s the proper way to spell it, not “wreckless” which if one thinks about is exactly the opposite). Since there were injuries there should be torts for all the injured parties plus Sound Transit’s insurer is going to come after the guy for indemnification.
Of course he’ll declare bankruptcy — unless he’s a multi-millionaire he certainly will be — so the insurance and all those injured will get only a small part of the damages levied. While most of the time I support bankruptcy as a way for people who have done no harm except misjudge their ability to repay debt, folks who have injured other people should be barred from it until they are made whole; the person at fault obviously has to have minimal housing and food, but everything else should go to the injured parties until they are made whole.
Given all the above, the at-fault driver’s life is going to be basically ruined. All his assets will be cleaned out and he’ll have to financially start anew from Square Zero. Yes, this crash is annoying, but really, that’s punishment enough. Stuff like this is bound to happen in any system that has grade crossings from time to tome.
So, you are excusing the driver who avoided the gates and trashed the LRV?
Joseph, your comment cracks me up, and illustrates what I mean about pitchforks. Which part of this quote from David B sounds excusing to you? What punishment is enough for you, or would be sufficient to satisfy the STB Commentariat Bloodthirst?
“the at-fault driver’s life is going to be basically ruined. All his assets will be cleaned out and he’ll have to financially start anew from Square Zero. Yes, this crash is annoying, but really, that’s punishment enough.”
And what about someone on the LRV receiving health care from the ACA and who has ongoing medical bills? After Trump and the Republicans cancel his or her insurance and Mr. Leadfoot declares bankruptcy, that person’s life will be ruined, and that person didn’t even run around a clearly marked crossing gate.
I’m not saying that there is such a person injured in such a chronic way in this accident, but deadbeat people take bankruptcy to avoid paying damages frequently. If the guy is single and doesn’t own a home, he’ll basically lose nothing after filing for bankruptcy.
I was driving in the tunnel when this occurred. The light rail operator initially had no idea what had happened, saying “I’m seeing faults and alarms…” while I could hear the alarms in the background of the transmission. He said that it appeared that somebody had activated the emergency brake and opened a door. LCC told him to take his hand radio to see what was going on, and to secure the opened door. Shortly afterwards, the operator reported that there had been a collision. Control asked if there were injuries, and the operator said that “[garbled] still alive”. Control assured him that aid was en route. My estimate is that it took over ten minutes for the first supervisor to arrive on scene.
They don’t have enough cameras or mirrors to see this from the drivers area?
If the SUV were sticking out of the side of the train, then yes. It looks like it hit and bounced off, so there wasn’t anything to see from the outside. You would have had to be looking that direction in order to see it happen.
Thanks for some first-hand information, Driver. Because details of this incident are something everybody interested in transit needs to know.
It’s good to know that one of our trains can take a collision like that and stay on the rails. I take it that the operator brought the train to full stop- at that speed “immediately” doesn’t happen.
Not good that it took a supervisor ten minutes to arrive a few blocks from Ryerson Base. Bus-side, Rail-side, I don’t want to hear it. Wonder how long it took Emergency Medical to get there. Also, any of the police.
And coldly, whatever the courts do with the motorist, I’d like to have a talk with them as to what was in their own mind when they drove around that gate. Most important thing to know as we design for prevention.
Mark Dublin
Would it be too much to hold off on condemning and jailing the driver until we actually know what happened? This is a bloodthirsty group when it comes to auto drivers!
What possible excuse could one have for trying to go around the gate and broadside a moving train? There’s no way you can do that without being negligent.
People have heart attacks, strokes or other medical conditions that cause them to lose control of their vehicle. Vehicles can malfunction. There’s surely a good chance of negligence, but I’m willing to keep my pitchfork at home until we know more.
I had a friend that would have hypoglycemic episodes due to his poorly controlled diabetes and do stupid things while driving. Fortunately for him the cops in Redmond recognized that he was having medical issues and didn’t go all the down their mental flow chart to the “shoot the uncooperative driver” step and called medics for him once they realized he wasn’t doing right.
If he were still alive I could totally see him running into a train.
R,
If he were still alive he should not have a license.
Ken,
Did both the throttle stick and the steering wheel break in such a way that the vehicle swerved around the lowered gate simultaneously. If so, the Tahoe needs to be hauled into court for attempted manslaughter.
Now there may have been mitigating circumstances; the driver may have refused to change the oil for so long that the poor, abused SUV vowed to take him down with it.
For someone with that kind of medical condition that isn’t clearly under control, the negligent act is getting behind the wheel in the first place. I would never dream of driving if I knew that kind of incident were a real possibility.
… though I’ve heard one or two stories of people who didn’t know they had such a condition until they got behind the wheel and found out the hard way.
In that very unlikely scenario, I would gladly retract my anger – though I still think his license should be taken away.
LRVs are tough. This much damage (together with the extensive damage to the SUV itself) is a clear sign that the driver was traveling much faster than the 30 mph speed limit on Holgate–I’d guess from the damage twice the limit or more. There’s no circumstance under which that’s OK.
Good point, Ken. It could be medical, it could be something else, we mustn’t assume he was just trying to save a couple minutes until we know.
I don’t mean to minimize the incident, but this isn’t even major news in most of the area news sources. I can’t even find Seattle Times coverage of it. The Stranger has an item, but no story. And then the TV stations all seem to have a story, but not the top headline.
Which is to say, t-bones happen frequently in the city, just not usually at a well marked train crossing. I don’t think this will provide political capitol to build a billion dollar tunnel. Maybe improved gates.
This was the first time a Link Light Rail train got t-boned.
Service was disrupted, and could have been disrupted for a lot longer. We reported.
Nor was this the first time we reported something the Seattle Times did not.
My comment was addressed to the commentators in the several lengthy threads above discussing grade separation and alternate routes.
I’ll fuck off.
Sincere thanks to the ST maintenance crews working through the night! I took the train this morning and it slowed to 5-10mph for the Holgate crossing, presumably due to the preliminary nature of any repairs.
I’m baffled by the collective shock and outrage. Everyone should know by now that you get what you pay for.
Typical motorist.
I’m sure he had somewhere really important to get to. Worth it!
Good observation, Ken. Andrew, explanations I can think of. Life-or-death (at least in the motorist’s mind) trip to hospital. Or gun pointed at the driver’s head. Or (again driver’s estimate) of deadly pursuit.
Again, like with any item in a major accident investigation, critical to know. Anywhere moving trains can suddenly occupy same space as moving automobiles, the physical and mental condition of thousands of individual motorists can cost or save lives.
And fletc(3)er, could be I’m still in toxic shock over the way our country’s news media handled this last Presidential campaign, and results of their performance. So while I think the SUV driver’s mindset is critical to know, news media has many other reasons to ignore besides its first reports of anything being invariably wrong.
But nobody peripherally mentioning public transit had better think that any collision on this order “happens”. Because that attitude is a hundred percent causal factor of next indentical one. Wouldn’t worry about “undergrounding” the E-3 Transitway. Whole area is water with a little dirt stirred in, so a billion won’t pay half.
Pillars above Stadium Station prove we can elevate. So next time you’re riding LINK approaching Holgate, if you see a gasoline tanker rolling toward those crossing gates at right angles, do quick cost-benefit analysis. Good thing imaginary ST- ballots are secret.
Mark
Christopher, I think the reason Central LINK chose Rainier Valley first is same as decision to start service with dual power buses and progressing through joint operations.
Given terrain, existing development, population and tax base, idea was to start with as much service as possible where most needed, and build from there. Always compared to a classic technical drawing: light hard pencil line, thickened and darkened over time.
Rainier Valley had many more regular passengers. Always some argument whether the 194 Express should have been kept in the Tunnel. But I think that Airport passengers would’ve massively favored train rides. Especially with luggage.
But I also think most cities with similar measurements and conditions would have had their first light rail line head straight for the airport. As will likely be included in next ST-, or the one after. Doubt it will leave from Westlake Station.
Some perspective. DSTT opened 27 years ago this September 15. Planning went back to at least 1983. I believe regional population was something like 1.5 million. Four million now. In the last three years, all our freeways have gone from jammed to impassable every rush hour. For truly regional distances.
Don’t think it’s complete accident that new figures and facts coincide so closely with opening of UW Station. As living creatures grow, so do their circulatory systems. Which, incidentally and unfortunately for rational transit development, never grow before the rest of the creature.
Mark
“I believe regional population was something like 1.5 million.”
And the commuter belt was Lynnwood – Bothell – Redmond – Renton – Kent – Des Moines. Everything beyond that was a separate job market. Except for Boeing workers who were often transferred between plants.
Rainier Valley was chosen partly because of the residential population and planned urban villages there (light rail is supposed to serve the people), and partly because serving a low-income minority area made it more favorable for federal grants. That’s another thing that has changed since then, the low-income part.
No offense meant, fletc(3)er. This medium is tone-deaf and expressionless. Also very hard to connect reply and original statement.
My target was really the ideas themselves, repeated year after year by people Elected and Appointed to Know Better. They even disrespect the Chief of the Highway Patrol about motorcycles in bus lanes!
Can’t blame Charles Dickens that the Ghost of Crashes Yet To Come never haunts decision makers’ Christmases. Very little on-site inspection from funding sources. One more reason I keep advocating express transit to Olympia.
Since his ORCA card would always have goose-quilled out for travel time and expenses, Ebenezer Scrooge would’ve had the ink dried on the parchment to finance the elevated two minutes after he saw the truck through his train window.
Often used to be tempted to hack your suggested course of action into last sentence of an Executive Summary.
Mark
Report on King5 News, including interview with eyewitness bus driver: http://www.king5.com/news/local/major-collision-on-holgate-st-in-seattle/411507181
On streets like Holgate and Graham, I’d like to see ST experiment with raisable, tough, indestructible bollards. They would at least prevent rogue vehicle from even approaching an oncoming train. The line along MLK will never be raised, it will never be tunneled. Cross streets will never be lowered under the tracks. If nothing else, imagine the hue and cry of local residents bemoaning years more construction along the route. A Rainier bypass as an express route to the airport will never be built. All of those projects are wishful thinking and would cost billions and we don’t have now and won’t have in the future that kind of money. In a “normal” society, that SUV would not have rammed the train–the driver would have patiently waited until the train passed. But these are not normal times. Lucky no passengers were seriously injured or killed. And the renegade driver just might get a slap on the wrist, nothing more. Preventing car/train collisions with a simple, low-cost technology like bollards could be a solution. Won’t do anything for suicide-by-train wannabe’s or distracted pedestrians, but then again, a body will neither harm a Link vehicle nor injure on-board passengers. Why not at least try it at one or two busy intersections?
Brookdale,
A “bypass” to the airport would cost about $400 million, not “billions”. Fully 2/3 of it could run on the ground between the existing tracks and I-5. Yes, there are some places where the hill would have to be excavated, but that’s not nosebleed work like tunneling through city infrastructure.
The north end junction is already there in the form of the access loop at the Maintenance Facility. Cross Airport Way north of Spokane and drop down to ground level under the freeway in the old railroad spur ROW that already exists, head south in it and you’re not exactly “home free” but running on some fairly useful ground. There are some tight spots where the LRV track would have to be elevated above the freight, like the Cleveland Heavy Rail line, but they are few and reasonably short.
Make the thing single-track in most places with “meeting” sidings and use it for “express” service to and from South King County.
It’s people like that who make me not want to drive. They get in a hurry, do dumb things like run into Link trains, break Sounder crossing gates, weave in and out of traffic on the highways and cause big accidents (seriously, like everyday — everyday there’s a blocking accident on at least one of our major highways). Why can’t they figure out how to get up 10 minutes earlier so they don’t have to be concerned about only going 65 MPH on a 60 MPH highway?
Grade separation would be nice. Someone mentioned bollards. That is a great idea. We should put in bollards at all the at-grade crossings. Bollards may have saved someone’s life if this had happened at 8 am instead of 8 pm.