Planes, trains, and yes, more trains

I’ve taken a train maybe a dozen times in my life. On the few occasions it happened, it was usually the most convenient, economical, or timely way to get somewhere, and every single time it was uneventful, pleasant, and completely forgettable. Trains have just never been my first choice. If distance is the deciding factor, I’ll fly. If flexibility, freedom, and enjoyment are the key factors, I’ll drive. And yes, the title of this post is a nod to the movie, though our trip only delivered on two of the three modes of transportation: a few planes and far too many trains.

On our recent two-week trip to Europe, we deliberately chose to see as much countryside as possible, which meant leaning heavily on the rail network instead of flying between each city. All told, we booked ten trains over the trip: four from Venice to Lucerne, three from Lucerne to Munich, two from Munich to Paris, and one final leg from Paris to London. On paper, it looked like a reasonably well-thought-out plan; I have the hours on four different train company websites to prove it.  This includes occasionally padding connection times just to make sure a family of four with luggage in a strange country had time to get from platform 4 to platform 15. Getting to this final itinerary wasn’t simple either. Ferragosto (an Italian National Holiday) closed off some of the usual mid-August options out of Venice, and the direct Zurich corridor we’d hoped to use was completely sold out, so we ended up rerouting through Basel and Mannheim just to get from Lucerne to Munich. That’s a fair amount of planning before a single wheel ever turned.

In practice, this mode of travel turned into the most inconsistent, and at times genuinely nightmarish, part of our trip.

Of the original ten trains we booked, two were cancelled outright and had to be rebooked in the fourteen days before we left home, including one cancelled the weekend before departure. Of the ten trains we ended up taking, three were either late leaving or late arriving, and one of those delays turned into a missed connection, which meant hunting down the next available train and losing a few hours in a station that we hadn’t planned on losing anywhere. Depending on how you want to do the math, that’s somewhere between a thirty and fifty percent failure rate. I did that math standing on a platform in Mannheim, and it did not strike me as a recipe for success.

What made it more surprising was that this wasn’t some brand-new, unproven technology. Europe’s rail system has been running for over 150 years. If any transportation network on the planet has had enough time to work the kinks out, it’s this one. And yet here we were, rebooking cancelled trains less than a week before departure, and later sprinting through Swiss and German train stations trying not to miss a connection that had already been delayed once. It reinforced something I already knew about myself: I like driving. I like the control, and honestly, I like the freedom and simplicity of it too.

But the part that actually stuck with me wasn’t the trains themselves. It was the reminder that even a mature, well-run, heavily resourced system can still be brittle without enough slack built into it. Every leg of that route depended on the one before it showing up on time, and one late train in Mannheim doesn’t stay contained to that train alone. It ripples into the next connection, then the one after that, and three cities later it’s a missed connection. The whole thing is only as reliable as its most fragile handoff. Yes, I’m referencing something I’ve often talked about in Operations: the weakest link. That’s usually a design problem, not a staffing one, and it’s a pattern I see constantly in the businesses I work with too.

The instinct when something breaks is to add oversight, more checkpoints, more people watching the process. From a Lean or Continuous Improvement perspective, this has another name: waste. In my experience, in these situations, the actual fix is almost always a deliberate buffer between the steps that depend on each other, a bit of intentional slack.  Never plan on actual capacity (or throughput) being equal to the theoretical capacity; machines and people don’t work that way. But if that buffer is what keeps one delay from cascading into three, and keeps a customer from missing their connection altogether, it’s earning its keep.

None of this stays contained to the operation, either. A missed connection doesn’t just cost the company a seat; it costs the traveller a few hours they didn’t plan on losing, and a bit of trust that the system will do what it said it would do. Multiply that by however many passengers were on our missed connection alone, and a single delay stops looking like an isolated operations issue and starts looking like a customer experience problem. Businesses that manage their constraints well protect that experience on purpose. The ones that don’t just get lucky, until the clock strikes midnight.

This shows up everywhere in business, not just on a rail platform. Handoffs between departments, supply chains, project timelines where one slipped deadline quietly takes out three downstream ones. The tighter and more efficient a system looks on paper, with zero slack and everything timed to the minute, the more exposed it usually is the moment one piece of it doesn’t cooperate. And eventually, something doesn’t cooperate. There is something called “Murphy’s Law”, after all.

In specific work terms, Goldratt’s “Theory of Constraints” speaks well to this topic. The idea is simple: find the bottleneck, then align every other step in the process to its pace rather than letting them race ahead and pile up work in front of it. The buffer we talked about earlier is that subordination step in action, protecting the weakest link instead of pretending it isn’t there.  And then, yes, work to elevate the constraint.

If there’s a broader lesson here, it’s a bit of humility about how figured out any system really is, no matter how long it’s been around. Europe’s trains have had almost two hundred years to get this right, and they still haven’t fully engineered the failure points out of the system. So, the next time someone tells me flying cars or autonomous manned drones are just around the corner, running like clockwork from day one, I’ll believe it when I see it.

For now, I'll keep driving when I can, keep booking a little extra buffer time when I can't, and keep asking my clients a version of the same question I probably should have asked myself before this trip. Where in your system is the one delay that, if it happens, takes three other things down with it? Chances are you already know the answer; you've probably just been hoping it doesn't come up. If you want a second set of eyes on it, I'd genuinely enjoy that conversation.