 |
Re: San Rafael, Novato SMART rail warning devices malfunctioning
Author: mook
Date: 09-29-2016 - 21:18
I would speculate that light rail:
1) it isn't as light as you'd think, though FRA-compliant like SMART should be heavier. Ask anybody in a car or light truck that's had contact with one; the train wins.
2) most LR in the US (not all, but most) is articulated, 6 axles per car; and usually runs in at least 2-car trains so 12 axles or more. More points of contact with the rail make detection more reliable (see other stories about minimum light-engine consists on Real Railroads for similar reasons).
3) a key point: LR runs very frequently compared to conventional rail, usually 15 min. or less between trains (30 min. max). So the tracks stay clean.
4) another key point: on PRW, LR typically does not rely solely on the standard crossing circuits and predictor; they're more for backup should signaling be interrupted. Crossing gates, especially on high-volume lines (15 min. or less between trains), are often or usually tied into the signal system. It's not clear whether SMART is doing that with its signaling and PTC; it might not be worth the cost and effort with relatively low-volume service so they have to get the normal crossing circuits working right, not just working.
5) more of an issue north of Ignacio, but a mix of higher-speed passenger (>50mph) and low-speed freight (<30 mph) makes for a difficult timing situation with grade crossings. But it sounds from the articles like they're having more trouble in the south. Maybe we don't know enough just reading random news stories (very likely).
I do appreciate SMART apparently trying to get it right. Railroading is a bad place to use 'agile' development practices (as in software) where you don't even try to get it right before releasing. The public should not be (and apparently aren't relied on to be) beta testers for SMART, which is the right way to do it.
1) it isn't as light as you'd think, though FRA-compliant like SMART should be heavier. Ask anybody in a car or light truck that's had contact with one; the train wins.
2) most LR in the US (not all, but most) is articulated, 6 axles per car; and usually runs in at least 2-car trains so 12 axles or more. More points of contact with the rail make detection more reliable (see other stories about minimum light-engine consists on Real Railroads for similar reasons).
3) a key point: LR runs very frequently compared to conventional rail, usually 15 min. or less between trains (30 min. max). So the tracks stay clean.
4) another key point: on PRW, LR typically does not rely solely on the standard crossing circuits and predictor; they're more for backup should signaling be interrupted. Crossing gates, especially on high-volume lines (15 min. or less between trains), are often or usually tied into the signal system. It's not clear whether SMART is doing that with its signaling and PTC; it might not be worth the cost and effort with relatively low-volume service so they have to get the normal crossing circuits working right, not just working.
5) more of an issue north of Ignacio, but a mix of higher-speed passenger (>50mph) and low-speed freight (<30 mph) makes for a difficult timing situation with grade crossings. But it sounds from the articles like they're having more trouble in the south. Maybe we don't know enough just reading random news stories (very likely).
I do appreciate SMART apparently trying to get it right. Railroading is a bad place to use 'agile' development practices (as in software) where you don't even try to get it right before releasing. The public should not be (and apparently aren't relied on to be) beta testers for SMART, which is the right way to do it.
