An airport becomes the latest proving ground for factory-built data centres

DXN Limited told the market recently in August 2026 that it had won a roughly A$4.1 million contract to design and build a prefabricated edge data centre for Australia Pacific Airports (Melbourne), the operator of Melbourne Airport. On the same day the company disclosed a further A$1.6 million variation to its existing Globalstar telecommunications work, lifting the day’s tally of newly announced Business to about A$5.7 million. For a company valued at barely A$7 million a few months ago, that is a meaningful haul. More importantly, it pushes DXN (ASX:DXN) into a new vertical: aviation, where downtime is not an inconvenience but a safety and security problem.

 

What just happened

The airport contract covers the full job. DXN will design, engineer, manufacture, factory acceptance test, deliver, install and commission a complete facility, including critical power, cooling, fire detection and suppression. The build sits inside APAM’s Airport Development Plan ICT program, which supports latency-sensitive systems such as baggage handling and security screening that need computing on site rather than in a distant cloud region. Site testing and commissioning are targeted for the first half of 2027.

The win lands only weeks after DXN’s headline event of the year. Recently in June 2026 the company announced a binding A$8.8 million contract to Supply a 1.36 megawatt AI high-performance-computing modular data centre to an undisclosed, US-listed neo-cloud operator. The stock, which had opened low, closed sharply higher that day, up 590 per cent, on Volume near 107 million shares. Market Capitalisation jumped from roughly A$6.6 million to about A$46 million.

 

Why it matters

Two contracts, two different worlds. The US deal is DXN’s first genuine foothold in the AI infrastructure gold rush; the airport deal shows the same factory-built model solving a very Australian problem: adding critical capacity inside a live, space-constrained site without a multi-year construction program. Managing director Shalini Lagrutta framed prefabrication as “the faster route to new capacity, and the better-quality one, compared with a multi-year upgrade.” If DXN can turn a marquee airport reference into repeat work with other infrastructure operators, the aviation channel could prove more durable than a single AI pilot.

 

The numbers that matter

The financial base is still small, which cuts both ways. In the March quarter (Q3 FY26) DXN reported Revenue of A$3.7 million, up 47.5 per cent year on year, with modular Manufacturing contributing A$2.9 million. The order Backlog stood at A$10.4 million, of which A$7.8 million was modular, and management flagged a pipeline of 89 identified projects. Cash was A$2.0 million at quarter end, supported by positive Operating Cash Flow of A$0.8 million.

The recent AI contract changed the balance sheet. DXN raised A$7 million through a Placement priced at 13 cents a share, lifting pro forma cash to roughly A$8.6 million. The 1.36MW Facility is being manufactured at DXN’s Perth plant, using prefabricated modules with direct-to-chip liquid cooling and support for high-density GPU racks, with commissioning targeted before the end of 2026. DXN has now deployed more than 80 modules globally.

The eye-catching figure is the follow-on. DXN has said the pilot could open the door to a campus-scale program worth more than A$278 million (about US$200 million) over one to two years. That is not a signed order, and investors should treat it as an option, not a forecast.

 

The bigger industry picture

DXN is riding two structural waves. The first is the scramble for AI compute, where operators need power-dense, liquid-cooled capacity online in months, not years. Traditional data-centre construction cannot keep pace, which is why factory-built modules, tested before they leave the plant, are gaining traction. The second is edge computing: pushing smaller facilities close to where data is generated, whether an airport, a port or a telecommunications site. DXN’s earlier work spans cable landing stations, a South American contract worth about A$5.3 million and a A$2 million modular Job for DP World Australia. The airport deal extends that logic to aviation. Both waves reward speed and standardisation, precisely the attributes a modular manufacturer is built to sell.

 

The bull case

The bull case is straightforward. DXN has moved from selling into telco and logistics niches to landing a US AI contract and a mission-critical airport build within a single quarter. Revenue is growing at a brisk clip, the backlog is rising, and the company is now generating positive operating cash flow at the quarterly level. A strengthened Balance Sheet gives it room to deliver the pilots that could unlock much larger programs. If even a fraction of the mooted A$278 million campus opportunity converts, DXN’s current revenue base would be transformed.

 

The risks

The risks are equally clear. This remains a micro-cap that reported March-quarter revenue of A$3.7 million, and a single contract can swing its numbers dramatically. The share price has run hard on the promise of the AI deal, so expectations are elevated and any delivery stumble would be punished. The largest opportunity, the US campus program, is unsigned and depends on a successful pilot for a customer DXN has not named. Execution risk on a high-profile airport project is real; a delay or fault could dent the reputation the company is trying to build. And the group is still expanding across Asia-Pacific, including an Indonesian joint venture, which stretches management and capital.

 

What happens next

Watch for several near-term catalysts:

  • Commissioning of the 1.36MW US AI facility before the end of 2026, the key proof point.
  • Any conversion of the mooted A$278 million campus program into a signed order.
  • The FY26 full-year result, for revenue, backlog and cash trajectory.
  • Progress and first commercial output from the Indonesian joint venture, flagged for around August 2026.
  • Further edge or aviation contracts that would show the Melbourne Airport win is repeatable.

 

Conclusion

DXN has engineered a genuine inflection. Within three months it has gone from a forgotten micro-cap to a company holding a US AI data-centre contract, a marquee airport deal and a materially stronger cash position. The strategy, standardised, factory-built infrastructure delivered fast, is well matched to where Demand is heading in both AI and edge computing. Yet the gap between promise and proof is wide. The revenue base is still slender, the biggest prize is unsigned, and the market has already priced in a lot of optimism. The next two quarters, culminating in the US commissioning milestone, will show whether DXN can turn a remarkable run of announcements into durable earnings. Delivery, not deal flow, is now the story.