A Tale Of Two Engineering Jobs

My New Job Put Me On a Boat

Since the last time I posted, I started at a new job. Within a week I was out walking on an actual job site, shadowing someone who'd been doing this for much longer. By week 2 I got on a tiny boat to look at piles (giant steel legs holding up piers, which are found on all kinds of marine structures - berths, bridges, etc).

This is kind of a massive whiplash from my first engineering job, where I spent probably 2-3 days just doing various inductions, meeting people, and sitting around feeling awkward and ignored. My first 'real' site trip that wasn't just a tour organised for all of the interns involved me going up there alone because everyone at the office was too busy to take me, and then sort of sitting awkwardly in a corner again.

I suppose part of this is the oddly industrialised nature of procuring new interns and grads at a large corporation. There is a pretty much annual intake of interns and grads, and by the time I was sitting on the other side of it as someone who's working there and having interns thrust into the team, I was also not really making much of an effort to try to include them into anything, too worried about my own problems, and also trying to rein in my increasing cynicism in front of these bright-eyed university graduates.

On the flipside, I was hired by a small local firm with, as far as I can tell, relatively little turnover. They brought me on because their pipeline of work was getting too large for their existing team - the owner, my boss, had a very clear vision for how I'd fit in before even posting the job ad.

It was really nice to step into a place where people have consciously set up a desk just for me - not just a physical place to sit¹, but a real sense that I'm here to do something. There is space for me here. I had my place in the schedule, there were things I'm expected to do - by my second day on the boat I got promoted from 'sit there observing and occasionally pointing out defects visible from where I'm sitting' to 'help take photos to speed things up'.

Obviously, I think following someone out to do a job is much more likely to result in faster and more effective learning than "my manager straight up ignores me for 3 - 4 business days because he's in back-to-back meetings". Instant feedback is best for learning, and is particularly good for me because being idle at work tends to turn me into an anxious wreck.⁴

This experience has got me thinking about training and knowledge and upskilling at work. Specifically, tying into a relatively obscure bridge collapse I've read about some time ago.

Case Study: The Malahide Viaduct Collapse

I learned about this from the Brady Heywood podcast (which covers various engineering failures). Brady has also written up an article about it, which is a good summary.

The bridge collapsed because the asset owner failed to keep doing the necessary preventative maintenance, and failed to inspect the structure adequately against the damage mechanism (which was scour, a type of erosion that happens in moving water).

The thing is, there's been in a bridge in this spot since 1844, and these stone piers have been there since 1860. Scour had been a known problem for this specific structure for at least at century. In fact, when they did a big refurb in 1965, they designed the above-deck structure with maintenance access in mind² - but by 2009 not a single engineer in the team realised scour was even a possibility.

The organisation that owned the bridge and was responsible for its upkeep is Iarnród Éireann (IÉ), aka Irish Rail, owner and operator of the entire rail network in the Republic of Ireland. It actually managed to lose information that had been passed down for at least a century, all the way from some Victorian era engineer to the present day. Interestingly enough, plenty of people alive today know about scour, even scour on that specific structure - but they weren't in a position to make sure the maintenance got done.

Here's what the national railway safety regulator's Rail Accident Investigation Unit (RAIU) said in the official investigation report:

Underlying factors to the accident were:

• There was a loss of corporate memory when former Iarnród Éireann staff left the Division, which resulted in valuable information in the relation to the historic scouring and maintenance not being available to the staff in place at the time of the accident;
• There was a dearth of information in relation to the Malahide Viaduct due to Iarnród Éireanns failure to properly introduce their information asset management system;
• Iarnród Éireanns inadequate resourcing of Engineers for structural inspections to be carried out at the Malahide Viaduct;
• Iarnród Éireanns failure to meet all the requirements of their Structural Inspections Standard, I-STR-6510

The Bridge of the Mind

I've previously discussed the idea that all man-made engineering artifacts - airplanes, bridges, motorcycles - are firstly mental phenomena before they are physical things that exist, because entropy works against man-made structures. Your bridge will fail without maintenance, and therefore you must know what maintenance to do and when to do it.

The custodians of a bridge, from the position of a manager's desk, cannot personally weld the steel or go up there and replenish the rock armour by dumping rocks onto the pier footings. In fact, in our highly complex industrialised society, our manager is usually above such considerations - but these considerations have to exist in some form in the system or the bridge will fail.

Therefore the modern manager's job is to be a custodian of the system that retains the image of the bridge. This task is knowledge management and must be done primarily by managing knowledge workers, potentially with their own managerial powers over lower-ranked employees or contractors³.

In that list of failings that the RAIU found, it was all knowledge management:

• The people who knew about scour left, retired, or were let go, without passing the information down to people who remained
• IÉ failed to set up a functioning information asset management system - the paper-based one was incredibly disorganised (reports would disappear into drawers, never to be seen again - a 1997 report noting heavy scour on the pier that collapsed was found by staff cleaning out an office. Also, no one actually used the new digitised database they rolled out
• Inadequate resourcing - corners were being cut on training, staff did not have a lot of time to touch base with one another, and infrequent contact with teams from other divisions meant that sometimes practices could diverge fairly significantly
• Attempts had been made to issue/ enforce formal standards, which were not being followed, partly because they weren't written very well, possibly because the people writing them were so far away from the actual work that they couldn't produce anything useful

In order to maintain the concrete bridge, you must maintain the bridge of the mind, and if this isn't done the actual concrete bridge will fail eventually. Not overnight - some systems are fairly robust - but a historically extremely underfunded place like IÉ is a lot more brittle, because failures tend to compound.

Unfortunately, the bridge of the mind can be wrong in extremely non-obvious ways. Sometimes people just have a different bridge in mind to the one that they actually have custody of - in this case, the engineers believed their bridge wasn't susceptible to scour, when in reality it was.

Much as actual concrete bridges are gradually damaged by entropy and corrosion, mental bridges are often subject to attack by management structures which are hostile to truth.

Damage Mechanisms in Mental Structures

How to create organisations that systemically produce good work at scale is kind of an unsolved problem.

Organisations that systemically destroy the possibility of good work, however, are dime a dozen. I'll list a few features relevant to this case.

A flexible definition of "done"

In addition to the RAIU investigation report, there were two rounds of safety reviews done on IÉ as an organisation - on in 1998 and another in 2006. The later report actually goes through how the organisation has responded to actions and observations from the first review.

From the 2006 AD Little report "A review of railway safety and of the role and function of the Railway Safety Commission"

Having reviewed the latest progress report and associated documentation we have three concerns:

• The allocation of project status for some projects appears to be potentially misleading, for example:
• • SMS1 (developing personal performance contracts) is reported as ‘in place’, although there is over €2m budget outlined in the remainder of the Programme, and the report states that the plan for 2006 is ‘ongoing application’
• • SMS2 (Infrastructure Asset Management System) is also reported as being ‘in place’, although again there is over €0.5m outlined in the remainder of the budget, and an extensive plan of training marked for 2006

The database system referenced in SMS2 was not being used when the IÉ Safety Performance Manager said the action was complete to the Railway Safety Commission, the railways regulator. When the bridge collapsed three years later, it was still not being used.

To some extent, this stuff is also what the regulator allows. The RAIU found:

The RSC provided a letter from the RSC to IÉ's Safety Performance Manager, dated the 9th April 2008 which documented that at that time the RSC considered the Recommendation S3 (i.e. to develop a flood/scour management system) closed “based on the evidence provided” by IÉ at the regular progress tracking meetings.

This flood/scour management system didn't actually happen. It seems like as the regulator, the RSC decided to sign off on this safety actions based on word from the Safety Performance Manager of IÉ's "trust me bro" at a meeting which probably never had minutes.

When "done" doesn't actually have to be done, it's natural that nothing, in fact, actually gets done.

Management Decisions Are Final

From the 1998 report, "A Review of Railway Safety in Ireland":

Past policy edicts emanating from top CIE and IE management over a number of years have created an environment in which senior engineering managers sometimes ceased to bid for essential finance for infrastructure renewal on the basis that there were known to be no funds available. Coping with less than adequate resources became accepted as the norm.

This has led to a weakness in the IE planning process, in that certain managers have refrained from preparing long term investment and maintenance plans on the basis that future funding (including EU) is uncertain. This in turn means that CIE officers do not have the detailed and appropriate investment plans backed by engineering evidence to put to government departments for consideration.

Management imposes some version of reality - say, we can drill the Macondo Prospect on this schedule, or we have this amount of budget to keep the trains running. Nothing is allowed to contradict this reality.

Trapped between the facts on the ground and the policies of managers that do not have contact with the ground truth, something gives. That something is usually the bridge of the mind, distorted under the load of the cognitive dissonance.

Being Highly "Efficient"

It takes time to build and maintain the bridge of the mind - to understand the job you are meant to be doing, and to keep reconciling the bridges of the mind with the actual concrete bridges and update one's understanding of the work.

IÉ's management had not respected the complexity of this task. Engineers were wildly under-resourced:

From the RAIU investigation, it was found that between the AE and ADE 2, there were inspections to be carried out on an estimated 2,150 civil infrastructure assets (519 of these were bridges). Given that these 2,150 assets all require different frequencies of inspections, the RAIU have calculated that the AE and ADE 2 each have approximately 800 (including 130 bridge inspections) inspections to carry out each year.

Two people, 800 inspections a year. That's a bit more than 2 structures each if both engineers worked every single day, including weekends and public holidays - an absolutely insane ask.

Except inspections weren't even their only job!

The RAIU have reviewed the job descriptions for both the AE and ADE 2. In the case of the AE, the carrying out of inspections is only one task of twelve specified in the AE‟s job description. For ADE 2, the inspection of assets only forms one of the eighteen tasks specified in ADE 2's job description.

Constant effort is required to match your understanding of anything with the ground truth. When understanding diverges from the ground truth, disaster happens.

From Nine Rules For Managing Humans Managing Nuclear Reactors:

You must establish simple and direct means to find out what is going on in detail in the area of your responsibility. There are ways of doing this, but all involve constant drudgery. For this reason, most managers avoid keeping up with the details.

Instead they create “management information systems.” Gimmicks⁵ such as these merely demonstrate you are not able or willing to use the necessary effort on your job.

But, see the above point about closing out actions which haven't actually been completed - it turns out the "efficient" executives turn around and enforce their brand of "efficiency" on the rest of the organisation, effectively mandating that no one is given enough space to devote sufficient attention to their tasks.

Employees who can't meet the standards of the organisation's "efficiency" will unfortunately need to be let go. Can't have you drag down the Velocity by being slow because you're actually doing your job to a respectable standard!

The other effect of having "efficiency" forced on the workforce is that often, training standards slip significantly. Training is a time investment, and a significant one. Senior personnel under serious pressure may elect to do everything themselves rather than running their juniors through the work, because they can do it faster with less risk of rework - but this means juniors never get to do things, until the day they somehow end up abruptly having to fill those shoes.

Sometimes there is so much work that managers rush the training and onboarding process itself, like the RAIU investigation found:

The inspectors are required to complete the 'Civil Engineering Earthworks and Structures – Inspection Guidance' training course, for which the training notes are available to the inspectors. This training course had been designed to be carried out over a three day period. However, the course was subsequently cut to a period of one and a half days, and some elements of the course were skipped or only mentioned, meaning attendees were reliant on the training notes provided.

This does not bode well for the bridge of the mind. Much like actual bridges, the one in the mind is not built overnight and requires specific tools and scaffolding. Expertise is built by exposure and practice and guidance. Without the last one, junior engineers are effectively learning via trial and error, and in particularly bad systems, they don't even get to know when they've fucked up, which is completely and utterly terrifying for anything important.

In systems where errors could have a body count, I would strongly prefer a lot more mentoring and actual training.

Learning On the Job

My first foray into large corporate engineering thrust me into a position within a business, where the business has problems for me to solve, but I didn't know enough to help. Other people were too busy to help or offer support.

But more worrying was the people in charge. They were very good at giving the impression that they cared, but by the time I left, I was very sure that engineering judgement was always going to subservient to management. There were plenty of people who started off as engineers, went into management, and seemingly forgot everything they knew or practiced. The engineers who I thought were really competent and vocal about their opinions were notably nowhere near any real levers for decision making.

I think a good cue to get out is when you realise that you don't really trust the judgement of the people telling you what to do (and you don't have any way to sway them to your side with charisma or other soft power tactics).

I am in no position to give advice, so I'll just share what I did.

1. Notice that I was in way over my head, and that I had no ability to tell if I was making mistakes.
2. Consider the potential consequences of these mistakes and whether I was prepared to live with them. Be realistic and definitely chat with a therapist if you have an anxious personality like myself - but even after I did therapy to manage my anxiety, I determined that incompetence on my part really could get people killed and that I cannot deal with the fallout
3. At this point I had choices to make. I tried to get better, good enough to convince myself that I could do it. I researched, I read up on incidents and failures and investigations.
4. I tried requesting more support from a manager and then found out that the management structure was actively blocking the most senior engineer on the asset from mentoring me (which he regularly ignored when I asked - but a hostile management is a hostile management, and the odds of winning these fights in the long run are dismal).
5. I quit - both with the aim of recovering my mental health, and also with the aim of finding work that would have me working with actual mentoring.
6. I spent a few months writing, networking⁶, and then finally made myself apply for jobs.
7. Somehow I got a job which put me on a boat - and so far it's been good!

I have certain advantages, true. I don't have massive liabilities or dependents, so my cost of living is pretty cheap, and the corporate job which taught me very little engineering while wrecking my mental health paid insanely well.

But signalling a willingness to actually learn and actually giving a shit about doing work properly goes a very long way with the kinds of people you'd actually want to work with, and other than my poor mental health, there was no reason why I couldn't have done a lot of that networking while I was still employed.

Morning Routines, Before and After

Let me walk you through my morning routine for my old job, and my new one.

Before I quit, I wake up at 6.30am and turn off my alarm. My heart rate immediately spikes, so despite my best efforts and my fatigue from being up late the previous night (insomnia), I fail to get any more sleep, so I check my calendar to make sure I don't have any morning meetings. I get up for real at 7am and make myself some breakfast, which I actually quite like, while telling myself it's a work from home day so I don't have to rush. At 7.30am I eat my breakfast, catch up on my group chats, and try not to think about work. At 8.30am I realise I have to start work soon (in fact I'm technically late, but after I took stress leave in 2024 I had an arrangement with my manager to start late whenever I need to). I sit on the ground and I try to distract myself from the fact that I need to work. On my good days I journal for half an hour to an hour until I feel okay enough to start. On my bad days I spend an hour on my phone and spend even longer journalling. Sometimes it's 10am before I manage to bring myself to login to my laptop. I read my emails. I try to figure out what I'm supposed to be doing. Sometimes I lose the will to do this, and I end up on the phone again and I hate myself. I wonder why the fuck do I struggle with the simple task of sitting at my desk and staring into the laptop.

At my new job, my start times vary - 6.30am for site days, 7am for office reporting days, and I don't have work from home days. I wake up between 5am to 5.30am, feed the cat, and feed the human (me). If it's an early start I microwave something from my freezer; but I sometimes manage to make myself breakfast too.

On my late start days, I can take the time to water the plants and play with the cat a little before I leave. I'm usually in the car between 6am to 6.15am, and I chuck on some tunes for my 20 minute long drive. The morning sky is usually beautiful, and sometimes I get a glimpse of the Swan River. Traffic is normally great since I'm early.

I roll into the office and make myself tea, then open my laptop. I know exactly what I'm doing, and if I don't, typically I can ask - I'm rarely the first in the office. If it's a relatively repetitive task, I put on some music.

It turns out I can do an early start and a long drive to the office in the morning as long as I'm not afraid of the work - and it turns out perks like being allowed to start whenever and work from home mean nothing when I am absolutely terrified of the work. And what I realise now is that I was afraid of all the intensely political nature of the work, which in my opinion left the technical work (i.e, the actual life-and-death stuff) neglected. I felt like I had a pile of insurmountable problems that no amount of technical acumen could tackle, and I felt totally alone with it - not because the people I worked with were selfish, but because they were structurally barred from supporting me long-term.

I now spend about an hour a day commuting, which is about twenty more minutes than my old job, or one whole hour if I'm working from home. I'm up three hours before my partner. I work pretty hard once I'm there - but I'm not miserable because I just don't dread work anymore.

I don't think that my career experience is typical. I think as someone with a high GPA and a decent resume, which got me hired at a large corporation in a technical role, I was expected to do my time, work for a decade or so, and eventually retire as some sort of middle manager. If I was more grifter-inclined, I could have networked harder with managerial types, participated in office politicking, and maybe gotten to upper management.

But would I have actually improved as an engineer? I truly don't think I could have. I suspect that it might be impossible to really learn engineering if you're in a cost centre, rather than a profit centre (though I've never worked in a profit centre, so I don't really know).

I don't think I'm alone in observing that grifters and incompetents seem to hold the reins of power at just about any institution with any kind of power. This, it seems, is just the world we have inherited from the era of shareholder primacy, austerity and neoliberalism. I also have no proposed solutions to any of this.

Instead, I've made a choice, and that choice is to tend to my own garden. I will put myself on a trajectory where I'll get to do lots of different things and meet lots of different people. I'll make sure to do my work intentionally and diligently - I won't do things that I think are literally worthless or harmful. I'll really think about the whys, the hows, the whys of the hows, and then I'll lock in and deliver the goods.

This isn't gonna put me on a path to riches, and I'm definitely not going to be get famous. I may not even get much influence. But so far, this approach is bringing me peace. I'm gaining skills. I'm doing something truly useful that will help people. I don't feel any dread before work, these days, and I won't give that up.