Ever been told your late starts are “costing the company money”?

The price of human downtime pales in comparison to machine malfunctions. 

While “absenteeism” costs the Australian economy about $33 billion a year, unplanned downtime in production plants costs Fortune 500 companies roughly US$1.5 trillion.

In factories mass-producing FMCGs, out-of-commission conveyor belts or faulty fans can cost $39,000 an hour. That's a lot of Tim Tams.

In automotive plants, the hourly cost of a major fault can jump to $2 million.

If there ever was an expensive problem in a large market… this is it.

Founded by Jean-Phillipe Picard and Tim Cheung, Factory AI combines AI with operational and sensor data to predict when critical machines will fail, estimate their remaining lifespan, and recommend optimal maintenance times.

The outcome of ‘fixing before failure’ could save Aussie businesses a median of $350,000 per hour.

When machines break

Industry 4.0 strategies and the growing sophistication of automated systems mean, basically, there’s more stuff that can go wrong. 

A small digital misfire can shut down an entire production line. Factories are overextended, running at full capacity 24/7 to fulfil pent up demand from the pandemic (yep, still). 

When the machines stop, the paid workers stop, production stops, and revenue stops.

In Australia, manufacturers (like Factory AI customer Arnott’s) lose an estimated $100,000 a month in lost production time.

Inflation is also a huge factor. Losing an hour's oil production when a barrel’s trading at $115 dollars feels a bit worse than when it was trading at $30.

In food and beverage, failure to deliver means lost supply chain contracts, lost shelf space, even paid penalties. 

Halted production lines during the pandemic sent manufacturers’ total losses into the stratosphere. Four years on, they’re still spiralling. This is one of the most monolithic, mass-scale pandemic hangovers in existence… 

And most people - except maybe those given a year-long wait for their new electric car - are completely unaware.

From alert to action

Factory AI is all about predictive maintenance. 

Not over-maintenance (costly and wasteful in itself), and not reactive maintenance (waiting ‘til something breaks and needs an emergency fix).

“Some of our customers lose millions annually on unplanned downtime. Factory AI detects anomalies so they can proactively tackle asset issues - not react to them - before they escalate into critical issues that shut them down for hours at a time.” 

When things like drops in pressure, temperature spikes, or vibrations occur, Factory AI sends notifications to an Engineering Manager’s device. 

Current solutions (Movus, Augury, IFM, Senseye) tend to offer fragmented or highly technical diagnostics, or dedicated hardware that’s only compatible with certain assets. (Factory is built for the average engineer, requiring no technical expertise.)

But its biggest competitor by far is legacy processes/the “old ways” - regularly scheduled maintenance visits, or sites simply burying their heads in the sand, hoping things won’t break, and allowing assets to operate until they inevitably do. 

Factory AI is hardware-agnostic, and offers site-wide asset coverage - meaning it draws data from any operational source the site already has. The more data it ingests, the better its predictive models become, and the more money it saves customers. A site leader at one of Factory AI’s customers said:

“I've never seen a business actually bring the Internet of Things and Industry 4.0 ideals to life until Factory AI. We were impressed by their ability to unlock the hidden potential in our existing data.”

Meet the fixers

Tim’s first job was taking orders at his parents' restaurant, getting paid in spring rolls and prawn crackers. He watched his parents work relentlessly, learning early that when equipment breaks, everything stops - orders back up, customers get hangry, and money walks out the door. Years later at AWS, he saw the same story at industrial scale - entire factory lines shutting down because no one saw failure coming.

JP discovered through years in sales that selling to manufacturers isn't about fancy pitches or buzzword bingo. It's about solving real problems. 

When they met through Antler, their skills aligned perfectly: Tim builds, JP sells.

As JP puts it: "Tim is incredibly detail-oriented, often spotting important things that I might overlook." Tim sees the same strength in their differences:

"Working with JP is like being on a ship you know won't sink, even if you have a thousand leaks."

A trillion-dollar problem

The numbers are staggering. Unplanned downtime costs Fortune 500 companies roughly $1.5 trillion annually. In FMCG factories, each hour of downtime runs $39,000. In automotive plants, that jumps to a cool $2 mil per hour.

But only around 1% of Australian manufacturing plants currently use predictive maintenance solutions. 

The potential is enormous - fixing before failure could prevent 1.6 million hours of downtime and cut maintenance costs by over 40%.

Proof in production

Factory AI has already landed major Aussie enterprise customers, including Bega, Darrell Lea, Tassal, and Arnott’s. The iconic biscuit maker has prevented $300,000 in equipment failures using the platform.

Their early success comes from a solution that's genuinely different. While competitors offer fragmented or highly technical diagnostics, Factory AI built something for everyday engineers - no PhD required.

And unlike solutions that need specific hardware, Factory AI is hardware-agnostic, working with whatever sensors and systems a factory already has.

This approach has recently received major validation from Amazon, who named Factory AI as one of only three recommended companies globally for predictive maintenance.

Why we invested

The combination of a massive market problem, proven early traction, and founders with complementary skills made this an obvious investment. 

What's happening in manufacturing today reminds us of enterprise software 15 years ago - an industry ripe for digital transformation but needing solutions that work with legacy systems and real-world constraints. Factory AI bridges that gap.

When we see customers saving six figures from a single alert and Amazon recommending them to their customers, the opportunity becomes clear. 

This isn't just a good product - it's the right product at the right time for an industry that desperately needs it.