Industrial tech
Putting the machine in machine learning
Last updated: 6 Jun 2024
Market 101
Machine learning smarts are arriving on the factory floor to help with quality control, equipment maintenance, training and the automation of paperwork. For context, Europe’s plants are pretty smart already, full of sensors, robots, 3D printers and other high-tech machinery. All of this stuff creates piles of data that can be tapped for insights into how to speed things up or find bottlenecks, product defects, safety risks and other areas for improvement.
And it’s been a handsome year for those looking to mine manufacturing data and plug AI systems into factories. Germany’s Daedalus, founded by one of OpenAI’s first engineering hires, raised $21m in February to develop its vision for AI-controlled factories. Belgium’s Robovision inked a $42m deal in March to bring computer vision intelligence to industrial machinery, while Switzerland’s EthonAI raised a $16.5m Series A in May for software that predicts maintenance needs and catches quality issues early.
Further evidence of a swell in AI manufacturing can be seen in the $28.5m Series B for UK-founded Oden Technologies, now based in New York, in April, while Apple a month earlier acquired Canadian startup DarwinAI, which visually inspects components during the manufacturing process.
Big manufacturers including BMW, John Deere, Mercedes-Benz and Siemens are using so-called digital twin tech and AI analytics to identify and correct shortcomings in operations before a physical product is born, as well as during production. So-called predictive maintenance startups in Europe are suddenly experiencing VC love. Ai:omatic Solutions — based in Germany, as so many of these manufacturing startups are — raised €2m in January for a system that detects potential machine faults early and helps reduce downtime. Tomorrow Things, also German, raised €1.5m in March for a similar product.
At the same time, we should expect to see more digital assistants enter factories. Stuttgart-based Assemblio creates animated step-by-step assembly guides for workers — the company raised €2.1m in February. Startups are also mining factory data to develop new safety tools: Sweden’s Buddewise raised €3.5m in March for a system that keeps a logbook of events marked urgent in factories, like when a forklift travels into an area it shouldn’t.
If this all sounds like factories are turning into digital cathedrals, well, it’s not the full story. AI manufacturing is in the very early days — somewhere around ChatGPT 0.2 in the development phase. Most of the big plants churn out more data than they know how to manage, while inflation and interest rates are keeping upgrades to a minimum.
Outside of AI, the other big change for factories that can afford it is a new robot coworker or two — see our recent briefing on Europe’s droids. As for 3D printing, the one-time hot innovation, it’s tempting to conclude the hype has worn off. Things have not completely died — France’s 3Deus Dynamics raised €10m in May, for example, but the stagnating industry needs “the shock that only M&A can provide,” as a recent FT article put it.
Early stage market map
Key facts
$100bn
the value of precision parts sold annually in Germany1
28.5m
people are employed in manufacturing in Europe2
$7.6m
is the expected value of Europe’s predictive maintenance market by 2028; a CAGR growth of 27%3
Trends to watch
Baby steps for AI on the factory floor
Daedalus calls itself Europe’s most modern contract manufacturer, but what does it do? The company takes orders for extremely high-end pieces from industries such as medical devices, defence and semiconductors, each requiring custom components for their products.
Founder Jonas Schneider told Sifted in February that Daedalus’ technology will let workers communicate with its system via natural language. So you could prompt it with instructions like: “I need to make a hinge for this part that needs to be thin in these places.”
This sounds like potentially a big leap forward — and far away from the reality of most factories. But imagine creating a painstakingly-complicated silicon wafer only to discover afterwards it’s not usable. In theory, the Daedalus approach could avoid costly defects like this: the company's pattern-matching tech could apply previous knowledge to configure machines for every new part.
Drowning in data
Not to immediately pour cold water on AI manufacturing, but data generated by factory-floor equipment is usually not in a useful state for AI models.
Fewer than one quarter of respondents in a recent MIT Technology survey say all or most of the data their machines generate is suitable for existing machine learning tools. And the bigger the manufacturer, the greater the problem.
A lot of hardware only talks to IT systems in very limited ways. Without better data foundations, manufacturers won’t be able to achieve much with AI.
Struggles to stay competitive
There were only 30 manufacturing companies in the latest FT ranking of the 1,000 fastest-growing companies in Europe, putting the sector outside of the top 10. It tells us something about the state of the region’s factories.
Europe’s ability to produce goods at globally competitive prices has been hampered by crises like Covid-19, higher inflation and rising interest rates. Soaring energy prices in the wake of Russia’s invasion of Ukraine have harmed things too.
Suffice to say, many plants will lack the money to upgrade or replace antiquated equipment to accommodate machine learning anytime soon.
Startups tracked by Sifted
Sifted take
AI startups say they can help factories reap rewards from the data trove of sensors, robots and more. It won’t be a quick transition. Data cleaning is costly and challenging and plenty of manufacturing plants are struggling with everyday costs. But, it feels like a safe bet that AI systems will take on more precision manufacturing — a world where the upfront cost of snazzy new tech pales in comparison to the risk of producing defective high-end gear.
Rising stars
AI-driven, self-powered smart sensor for steam monitoring to increase energy efficiency, reduce maintenance costs and carbon footprint. The Turkish startup is backed by Springvest and Alesta Investment.
Round
Seed
Valuation
Undisclosed
Date
2022
Size
€10.9m
Beneficiaries of a huge grant investment from France 2030, the Parisienne startup designs and operates atomic thermal generators to massively decarbonize industrial emissions.
Round
Grant
Valuation
Undisclosed
Date
2023
Size
€32m
Ethon AI
Safety & quality and maintenance
Total funding
€7.2m
Backed by established investors Earlybird Venture Capital, Index Ventures, General Catalyst Partners and Founderful, Ethon AI aims to redefine quality management for manufacturing.
Round
Seed
Valuation
Undisclosed
Date
2023
Size
€6.2m
Outsight offers software solutions to track the motion of people and vehicles using 3D LiDAR data. It’ss backed by investors including Bpifrance, BNP Paribas, Groupe ADP and Energy Innovation Capital.
Round
Early VC
Valuation
Undisclosed
Date
2022
Size
€22m
Early stage startups to watch
3Deus Dynamics
Additive & parts manufacturing
€13.2m
€10m
-
Actuation Lab
Safety & quality and maintenance
€570k
-
-
ai-omatic
Safety & quality and maintenance
€2.7m
€2m
-
AiSupervision
Industrial systems
-
-
-
Bren
Safety & quality and maintenance
€11.2m
€10.9m
-
Buddywise
Safety & quality and maintenance
€4.9m
€3.5m
-
Carbon Instead
Sustainable solutions
-
-
-
Carbon Waters
Additive & parts manufacturing
€4.1m
€2m
-
ChemiTek
Safety & quality and maintenance
€2.3m
€1.2m
-
CNCTeile24
Marketplace
€8.3m
€8.3m
-
Continuum Industries
Developer & design tools
€14.2m
€9.1m
-
Cyclize GmbH
Sustainable solutions
€5m
€4.8m
-
Daedalus
Additive & parts manufacturing
€37.4m
€19.1m
-
Deltaray bv
Safety & quality and maintenance
€4.3m
€4.2m
-
DessIA Technologies
Developer & design tools
€5.5m
-
-
ENLYZE
Safety & quality and maintenance
€2.2m
-
-
EthonAI
Safety & quality and maintenance
€23m
€15.3m
-
gemineers
Safety & quality and maintenance
-
-
-
Headmade Materials
3D-printing
€12.3m
€8.3m
-
how.fm
Training tools
€7.5m
€4.9m
-
HT Materials Science
Additive & parts manufacturing
€20.6m
€13.6m
-
IPercept
Safety & quality and maintenance
€2.3m
€1.5m
-
Jimmy Energy
Sustainable solutions
€51m
€32m
€75m
Kognic
Industrial systems
€29.8m
€21.8m
-
Kraaft
Safety & quality and maintenance
€3.2m
€3.2m
-
Lekatech
Sustainable solutions
€1.7m
-
-
LIDROTEC
Lasers & imaging and sensors
€5.9m
€5m
-
Lithium Lasers
Lasers & imaging and sensors
€4.2m
€3.5m
-
MantiSpectra
Lasers & imaging and sensors
€4m
€4m
-
Mavarick
Sustainable solutions
€1.4m
€1.3m
-
MetisMotion
Industrial systems
€5.5m
€3m
-
Mindtrace.ai
Industrial systems
€4.1m
€2.6m
-
Miraex
Lasers & imaging and sensors
Quantum internet
€2.3m
€2.3m
-
Modelwise GmbH
Safety & quality and maintenance
€3m
€2m
-
MoonVision
Safety & quality and maintenance
€2.6m
€900k
-
MotionMiners
Lasers & imaging and sensors
€5.2m
€5m
-
Neural Concept
Developer & design tools
€9.9m
€8.3m
-
Operations1
Process optimisation & analytics
€15.3m
€11.1m
-
OTee
Industrial systems
€1.4m
€1.3m
-
Outsight
Lasers & imaging and sensors
€42.6m
€22m
-
Plyable
Developer & design tools
€3.5m
€2.9m
-
Qviro
Marketplace
€1m
€1m
-
remberg
Process optimisation & analytics
€13m
€11m
-
RESONIKS
Safety & quality and maintenance
€2.8m
€2.6m
-
Robok
Safety & quality and maintenance
€4.2m
€2.1m
-
Senorics
Lasers & imaging and sensors
€12.3m
€8m
-
SiteFlow Solution
Safety & quality and maintenance
€11.5m
€10m
-
Smart Steel Technologies
Process optimisation & analytics
€2.1m
-
-
Smartex
Safety & quality and maintenance
€23.7m
€22.4m
-
smartFAB
Process optimisation & analytics
€900k
€280k
-
SPREAD
Process optimisation & analytics
€17.1m
€14.6m
-
Synera
Developer & design tools
€19.7m
€14.5m
-
Tacto
Process optimisation & analytics
€55.3m
€50m
-
ToffeeX
3D-printing
€7.5m
€6m
-
Tomorrow Things
Sustainable solutions
€1.5m
€1.5m
-
Tvarit
Sustainable solutions
€14.3m
€6.5m
-
Upnano
3D-printing
-
-
-
Vanilla Steel
Process optimisation & analytics
€4.2m
€2.9m
-
Wayland Additive
Additive & parts manufacturing
€21.6m
€5m
-
Workerbase
Process optimisation & analytics
€13m
€10m
-
Xentara (by embedded ocean GmbH)
Process optimisation & analytics
€3.2m
€2m
-
Europe’s success stories
Who early stage startups are up against
(Pre-)Seed
Series A
Series B
Series C
Series D+
IPO/Exit
This company, which has banked almost $73m, has developed a wearable mobile mapping device that captures complete spatial data and 360° panoramic images of a construction site.
(Pre-)Seed
Series A
Series B
Series C
Series D+
IPO/Exit
German startup providing adaptable, affordable and connected full-stack robot automation solutions for repetitive manual worker tasks in industrial SMBs. It recently secured a $43m Series B round from Lightspeed Venture Partners, Sequoia Capital, Promus Ventures and Kindred Capital.
Your feedback
How would you rate this briefing?