A huge factor in successful software projects has nothing to do with code – although it is “soft”. It’s how you manage organisational structure. Building understanding between teams… aligning objectives as milestones are reached… making sure work keeps progressing and people stay looped amid complexity.
It’s hard work. And if you’re creating software for high-tech equipment – where coders face an additional challenge of the products all being new, innovative, and extremely expensive – the complexity skyrockets.
While AI is part of the answer, vibe-created code suffers accountability and responsibility issues that aren’t ideal for high-tech equipment software challenges. Which means many important judgements about what to do, how to do it, and how to apportion tasks and goals remain human-led.
In this blog, we’ll explore the Big Five issues that lead to slowdowns in commissioning high-tech equipment – the same challenges we’ve learned to navigate here at ICT Strypes over more than a decade working in high-tech system development for our clients. Let’s flip the On switch.
Early on: decisions made at the start echo down the years
On small-to-midsized projects – say, a new customer-facing corporate app – software development teams can build forks and spec creep into the way they work: they can stay nimble and agile, with application code often separated from underlying platforms.
Everyone knows Sales might want a new feature to win a €500,000 contract. Everyone knows programme goals can change or expand over a few months. So tooling and infrastructure are chosen with flexibility in mind.
But if the project is high-end semiconductor equipment – where machine control systems and embedded software are part of multi-year, billion-dollar rolling investment programmes – those options don’t exist. A modern chip fab or AI datacentre is a tightly integrated assembly with millions of moving parts. With some devices so niche they’re practically handmade.
In this world, early architecture decisions matter a lot more than in building a web app. Even if the code all works, there may be opportunity costs missed: long-term maintainability, how easily it can scale, time-to-market when a product cycle may be less than four years. Competitiveness may not mean “getting it wrong”. It may just mean “being sub-optimal.”
The way to combat this is to make the really big decisions a “contest”. Don’t simply choose a tooling pathway or tech stack you’ve used before; poll several options and pit them against each other, with criteria deciding the winner. With thousands of separate applications needed for a single fab production pathway, it’s wise to apply your keenest analysis to these fundamental choices.
People versus systems: complexity can scale even if people don’t
Big and complex hardware, controlled by big and complex software, means the engineering team may be big and complex, too. Which doesn’t so much triple the problem as cube it, raising it to the power of three.
One more IC on a motherboard, one extra feed in a power array, and the system isn’t just 2% harder to understand; it’s 10%, 50%, 2-3x. Because that single addition led to new capabilities to be defined, different performance envelopes to be explored, a thick stack of What-Ifs to be checked one-by-one. Which means longer build times, more drawn-out deployments, and a greater requirement for expertise.
So that’s our second point: once you’ve got your basic choices set – technologies, standards, practices – do the same for your team structure too. Reporting lines, org charts, the responsibility matrix, who takes decisions and who’s ultimately accountable are important. But also vital is that softer stuff: can they all communicate freely, work together in friendly co-existence, share bad news without reprisal?
This is how we do things in the ICT Strypes’ High Tech Team: treat the choice of people as just as vital as the choice of platforms and tooling. (Our Nearsurance methodology, for example, pairs a technical team at our development hub with a business team at the client’s own premises.) So look at your team with clear eyes – and a sense of nuance. If some people don’t suit the project environment, or others are working against their natural preferences, it’s time to reconfigure.
Bringing it all together: integration with a higher parts count
As fabs and factories get bigger, high-tech systems are integrating more and more subsystems and functionality under one roof. So machine control software today isn’t a list of apps; it’s an ecosystem, all the parts as intertwined as a rainforest.
Output from App 1 informs App 2; the operating range of App 3 limits what Apps 4, 5, and 6 can do. A server falling over with App 21 on it shuts down a precision process in App 54. And rolling upgrade programmes mean new apps get introduced all the time as legacy ones are deprecated.
In fact, successful software projects in high-tech systems today are largely a matter of how well you do this. Commissioning high-tech equipment is all about systems integration challenges and how you address them. Once again, it’s about optimising the connections between things as much as the things themselves.
We have long experience of this – creating software that controls the world’s most complex and precise hardware for producing the chips that make your phone smart, your servers of service, and your cloud always overhead. Among the (many) things we’ve learned is that systems thinking comes first.
Nothing exists in isolation; everything is part of an interconnected process. Treat them as such.
Personal differences: reconciling skillsets across organisations
Unlike semiconductors – where every chip is identical down to almost atomic resolution –you can’t stamp out people with a cookie-cutter. So scaling engineering teams isn’t as simple as assigning names from a list of developers.
Michaela’s brilliant with OpenDB, but her knowledge of datacentres makes her better suited to Client X than Client Y. Sam’s an expert with Kubernetes, but doesn’t he prefer C++ to Rust? And Monica, Martin, and Rafaela have all produced great results … but they all speak different languages.
Skills often live in silos, clustered in different offices or countries. Mix in different project cultures and ways of working – especially if you’re an international organisation with many offices – and you may find the skills available to you aren’t as extensive as you thought. And today’s embedded software and system integration challenges aren’t forgiving of everyone going their own way.
As our solution, what we look for in a team is: multidisciplinary first. ICT Strypes people come from a wide variety of professional experiences: some are CompSci PhDs, some dropped out of High School, some even have liberal arts and humanities backgrounds. But where they all score high is the ability to work across silos. Skills like a new language or runtime can be learned – but multidisciplinary is an attitude. An increasingly vital one.
Software will eat everything: today’s system of the world
In past decades, software was plumbing: it connected datasets and applications, and as long as it worked nobody thought much about it. But for every high-tech enterprise today – and plenty of low-tech ones – it’s become far more.
Software is now a competitive advantage. It doesn’t just pass data around; it transforms the processes it’s part of. Optimising pathways, improving workflows, driving down costs and boosting profit margins. Even – in the Age of AI – dreaming up new deterministic control systems and software architectures for machines, and executing them itself.
Software is becoming a critical driver of system performance and delivery timelines, not just a supporting function. Which puts our first four issues in stark perspective: the development team isn’t just answering a project brief – it’s contributing to the future health of its high-tech client.
Thankfully, this point is perhaps where ICT Strypes is strongest. Working as a team for our client’s future success is deep in our DNA – because that’s how ICT Strypes evolved, coming into existence as part of probably the highest-tech venture on Earth. To be fair, it’s a very difficult skill to learn … but we were born with it. Which gives us an advantage.
CONCLUSION: Complex high-tech systems integration starts with ICT Strypes
Among all software challenges, the high-tech equipment sector probably has the biggest. But we wouldn’t want it any other way. The production processes our software enables grow the world economy, empower billions of people, open up new sources of competitiveness every day. And the harder the challenge, the greater the opportunity.
So if you recognise yourself in the issues above – no matter which high-tech organisation you’re part of – book an initial call with ICT Strypes’ High Tech Team. If you own one of these challenges, we’ll own it with you.