Alexander van der Steeg 🚀

Gewinne Kontrolle und Ordnung über Deine IT-Assets zurück!

Ich bin ein technischer Enthusiast und Lösungsfinder, der seit vielen Jahren versucht mit Technologie das Leben zu verbessern.

Durch meinen Werdegang und der offenen Art habe ich in rund 20 Jahren sehr viele Branchen, deren Probleme und Werte, kennengelernt und dort jeweils mein Glück gefunden.

NEURA Robotics

Reporting directly to the CGO, I serve as the Growth Strategy Manager at NEURA Robotics. In this key position, my main goal is to identify and implement strategies to drive business expansion and market penetration.

I support the Chief Growth Officer by offering strategic advice, overseeing cross-functional initiatives, and empowering sub-departments such as Sales, Marketing, Service and Support, and People and Culture to achieve their objectives, and optimize operational efficiency.

NEURA Robotics (founded in 2019 near Stuttgart), is a German deep-tech company with a diverse team working on revolutionary robotics technologies to bring robots and humans closer together under the guiding principle "we serve humanity."

Growth Strategy Manager

Micropsi Industries

As a Deployment Strategist, we collaborate with a team of Engineers, Product Designers, and Sales Reps to deploy our software, MIRAI. We analyze major client challenges, interpret data, and identify product needs and user motivations for optimal impact.

We engage deeply with customer workflows, working with our customer success teams to understand data and product intricacies, aiding scalability.

Head of Deployment Strategy

As the inaugural Customer Success Manager, the unique opportunity to build and shape the Customer Success Department from scratch, focusing on creating effective strategies and fostering strong client relationships was something I really enjoyed. My role also involved productive collaborations with the engineering and product teams.

Artificial intelligence, software, software agents, robotics, robot control software, process prediction, predictive maintenance, and IoT.

CS Manager

GlassDollar

In my role as a Corporate Innovation Manager, I had the privilege of driving forward-thinking initiatives and fostering innovation across various projects with industry leaders like Daimler, Porsche, and BSH. 

My focus was on bridging the gap between traditional practices and innovative solutions, ensuring a seamless integration of cutting-edge ideas into established corporate frameworks.

Corporate Innovation Manager

Siemens Energy

Gas Turbines - Project Management: Innovation Lab

Collaborativ Robot (UR5) Implementation in Combustion Manufacturing.
In charge of planning, coordination and internal financing. Programming of the Cobot in the CAD / CAM environment (NX18), as well as process adaptation in the real working environment.

Digitalization/Paperless Manufacturing - Test reports as a cloud-based app
User-centred prototyping: record user specifications, describing user stories / jouneys, furthermore design, test and validate digital mock-ups (UX/UI with Adobe XD) with users.

Innovation Lab Project Driver

Siemens

Gas Turbines - Project Management: Innovation Lab
Within a Start-up enviroment, the manufacturing team creates new processes and products by indepth knowledge and out-of-the-box-thinking, agile working and fast prototyping (3D printing).

Responsible for the Innovation Lab - PCS / ULN Burner Support Housing.
Leading a team (>10 team members) by planning and monitoring the implementation of the technologies developed by the project. Contributing with the conceptualization, coordinating multi-disciplinary communication and business evaluation.

Innovation Lab Manager

Gas Turbines - Project Lead Factory Planning: Transformation
Transforming the manufacturing operations into modern working cultures, i.e. self-organizing teams, agile working modes and decentralized autonomy. 

Assistance in drawing up business plans, forecasts and budgets. Together with technical experts the development of business presentations for internal investments: cool bearings, housing and rotating components.

Project Management Assistant

Gas Turbines - Combustion Manufacturing: Apparatuses
Cataloguing and prioritisation of assembly fixtures, creating an maintenance and repair catalog and associated with it, developing and implementing an homogeneous concept for apparatuses.

Co-Project Manager Manufacturing / Production Maintenance

Garden Center Rudow - Tree Nursery

E-Commerce. Sales and customer service, coordinating supply logistics, work and process planning, assigning and scheduling of temporary employees.

Nurseryman

GermanyConcierge Incoming & Events GmbH

Event management, booking and sale, reservations, recommendations and arranging for services. Customer care in an international, high-end environment.

Concierge

Wilking & Tarrach Metallbau

Design/Concept and planning, manufacturing of construction elements, supporting structures and general metal processing. Schedule planning and client communication.

Intern

Der Holländer Pflanzencenter GmbH

Sales and customer service, coordinating supply logistics, work and process planning.
Consultant and Salesman

Robotics & AI: simplified | 44.000+ Followers | @micropsi | Ex-Siemens

I’m deeply interested in deep tech startups: 
building, bringing them to market, and scaling. Follow along as I share my journey and what I’m learning along the way.

Ilir Aliu

You need $100 Billion to solve self-driving? Wrong!
A small team in San Diego is doing it with a smartphone chip and a few cameras... for $999. 🧵

Waymo spent $10 Billion. comma.ai spent $18 Million. Who won?

If you want to understand capital efficiency, look at self-driving cars.

The industry is famous for burning cash. Cruise, Zoox, and Waymo have spent a combined $20 Billion+ trying to solve autonomy.

They built "Robotaxis"... $150,000 vehicles packed with LIDAR, geo-fenced to tiny areas, and terrified of rain.

Then there is George Hotz and comma.ai.

Hotz (the hacker who jailbroke the iPhone) took a different bet: Don't build the car. Fix the car people already own.

He built a $999 aftermarket device. You stick it on your windshield. Plug it in. And suddenly, your 2016 Honda Civic has better autonomy than a 2024 Mercedes.

The industry laughed. Then the regulators sent cease-and-desist letters. So Hotz did the most "hacker" thing possible: He open-sourced the code.

Fast forward to today:
- 20,000+ active users.
- 300 Million miles driven.
- #1 Rank from Consumer Reports (beating Tesla and Ford).
- Profitable.

While Big Auto builds "walled gardens," Comma is building the "Android of self-driving." They prove that you don't need billions of dollars to solve hard tech problems. 

You need better code, a community of 1,000+ contributors, and the courage to ignore the "rules" of the industry.

We are watching the Linux moment for robotics happen in real-time.

The "Stakes" are massive. This is a battle for the soul of AI.

Option A: AI is owned by 3 mega-corps who rent it to you for $15k. 
Option B: AI is open-source, runs on consumer hardware, and you own it.

comma.ai  is the only one fighting for Option B.

If you are tired of monthly subscriptions for heated seats, check them out.


Listen to my full interview with CTO Harald Schäfer here:
https://lnkd.in/dsdmM2rC

AI & Robotics | 100k+ | Scaling Deep Tech

If you work with robotics, AV, or 3D vision, this update will save you months of engineering.

Most models need complex engineering to get reliable 3D geometry.
This one does it with a plain transformer.

Depth Anything 3 is the new model from ByteDance that predicts stable, spatially consistent geometry from almost any visual input. No special architecture. No multi-task setup. Just a single transformer trained on depth rays.

✅ Single vanilla transformer as backbone
✅ One depth ray target instead of multi-task learning
✅ Outperforms DA2 in monocular depth
✅ Beats VGGT across camera pose, any-view geometry, and visual rendering
✅ Trained only on public academic datasets

It shows that simple designs can scale further than expected.
Cleaner pipelines. Fewer moving parts. Stronger results.

Project page: depth-anything-3.github.io



—-
Weekly robotics and AI insights. Subscribe free: scalingdeep.tech

Robot models get better only when humans feed them more demos. This one improves by learning from its own mistakes.

pi*0.6 is a new VLA from Physical Intelligence, that can refine its skills through real-world RL, not just teleop data. 

The team calls the method Recap, and from what I can see, the gains are not small.

A quick summary:

✅ Learns from its own rollouts using a value function trained across all data
✅ Humans only step in when the robot is about to drift too far
✅ Every correction updates the model and improves future rollouts
✅ Works across real tasks like espresso prep, laundry, and box assembly
✅ Throughput more than doubles on hard tasks, with far fewer failure cases

What stands out is the structure: 

a general policy, a shared value function, and a loop where the robot collects data, improves the critic, then improves itself again. 

No huge fleets of teleoperators. 
No massive manual resets.

If VLAs can reliably self-improve in the real world, the bottleneck shifts.

Data becomes cheaper. Deployment becomes the real test bench.

Full paper, videos, and method details here:
https://lnkd.in/dgCeZdjT

Most humanoid projects talk about real work.
Very few last an hour on a real line.
This week I saw a case that matters for anyone building robots, perception, or physical AI.
Kinisi deployed its first mobile manipulation system into a live recycling facility. Not a demo. Not a staged test. A real production line with real output pressure. 

Why this matters if you want robotics to deliver real value on your floor:

• Handles mixed glass with random poses and no fixed fixtures.
• Runs real grasp selection under noise, vibration and production variability.
• Maintains throughput while avoiding breakage on a delicate material.
• Shows mobile manipulation doing actual shift work instead of controlled lab runs.

Kinisi published a video that shows what the robot sees and how sensor data turns into action. This is the part most teams struggle to explain to customers, so the educational angle is useful for anyone working on adoption. 

On top of this, the team signed a pilot with a global automotive manufacturer to explore humanoid use cases in production. 

The direction is clear. 

Wheeled mobility (not legs!) plus strong perception seems to be shaping a large part of industrial humanoids right now. 

I know Brennand from earlier conversations and from our podcast session, and I am always glad to see European teams push the category forward. Wishing the Kinisi team continued success.




—-

Weekly robotics and AI insights.
Subscribe free: http://scalingdeep.tech

A simple CNC machine. Wood. A rotating tool. And yet the pattern shifts from circles to perfect hexagons with a level of accuracy most factories still struggle to hit.

I worked five years in manufacturing at Siemens, and this is the part people often overlook.

The real magic is not new or loud.

It is stability. 

Repeatability. 

Tool paths that were tuned forever. Precision that just keeps working.

Robotics is in a hype cycle right now, but manufacturing has delivered quiet perfection long before anyone talked about foundation models.

Sometimes the most interesting engineering is the part that does not ask for attention.



—-

Weekly robotics and AI insights.
Subscribe free: scalingdeep.tech

Humanoids learn dexterity from videos? Wrong!
Make a fist, cameras go blind. A tiny Dutch startup fixed what billion-dollar labs couldn’t. 🧵

In 2014, Stephan van den Brink was a tax consultant at Ernst & Young. He had a stable career in fiscal law. But he couldn't stop thinking about Virtual Reality.

He saw a gap: "We have headsets to see, but no way to touch."

So he quit. He founded MANUS™ to build the ultimate VR gaming glove. They flew to E3 in Los Angeles. They slept in cheap motels, and snuck their booth equipment past security because they couldn't afford the union fees.

They built a sleek, futuristic glove for gamers. And nobody bought it.

Gamers didn't want to wear gloves. It was too much friction. Most startups would have died there. Manus didn't. They pivoted.

Stephan and his team realized that while gamers didn't care about finger precision, engineers did. 

• NASA needed it for simulation. 
• Hollywood needed it for motion capture. 
• And eventually... Robotics needed it for AI.

Today, the Humanoid Robot revolution is built on data. 

To teach a robot how to fold a shirt or crack an egg, a human has to do it first (Teleoperation). Cameras (Computer Vision) often fail because fingers block each other ("occlusion").

Manus had spent 10 years perfecting Electro Magnetic Field (EMF) tracking. It doesn't use cameras. It uses magnetic fields to know exactly where a finger is, with millimeter precision, even in the dark.

When Tesla and other giants started building humanoids, they didn't build their own gloves. They called Manus.

I had Stephan on the podcast to discuss this incredible 10-year journey. We talked about:

The hustle, how to launch a hardware startup with no money, on how to survive when your target market (gamers) rejects you, why magnetic fields beat computer vision for ground-truth data, and last but not least:

How human dexterity is being "uploaded" into robots.

This is a masterclass in resilience and finding "Product-Market Fit" in the most unexpected places.

Listen to the episode here:
https://lnkd.in/dD68NUqX

And thanks to Dove Feng for making this happen!

Most robotics startups never make it out of the lab.
This one raised $20M and deployed in 100+ factories.
[If you build robotics, save this 🧵]

This is one of the clearest lessons I’ve seen on execution vs technology.
[All lessons summarised at the end 👇)

Three founders (Rudy Cohen, Louis DUMAS, and Albane Dersy) built a factory-first robotics company now deployed across Europe, the US, and Japan.

They had:

• No hardware background
• No manufacturing network
• No industrial reputation

But they raised $20M and deployed in 100+ factories by doing what most founders avoid:

They treated execution, not tech, as the product.

Most deep-tech teams make the same mistakes:

• Talk to 5 customers, not 50
• Polish the product for years
• Expand only when a country “is ready”
• Treat deployment as a support function

Inbolt did the opposite.

Here’s where it started.

Not in a lab.
Not in a research paper.

But in a bus maintenance shop, watching workers struggle with daily tasks.
Real pain.
Real chaos.
Real constraints.

Since then, the three founders have spent more time on factory floors than most executives will in an entire career.

And the magic was the mix.

• Rudy with deep engineering roots
• Louis with aerospace and robotics expertise
• Albane with operations, sales, and commercial execution

That combination is what made the company work.

The execution lessons:

1. Talk to dozens of customers.
Not five.
Factories are fragmented.
You need operators, maintenance, quality, integrators… all of them.

2. Sell early.
They closed deals before the company was even fully set up.

3. Pivot based on customer pull.
Their early idea guided human workers.
Customers said, “Put this on robots.”
So they did.

4. Pair technical excellence with operational excellence.
The tech handles unstructured environments.
Albane built the systems to deploy it fast and reliably.

5. Deployment is a strategic weapon.
Most teams outsource it.
They didn’t.

6. Go global early.
Not because it’s cool.
Because automotive OEMs operate everywhere.

7. ROI > AI.
Factories don’t care about “intelligence.”
They care about uptime, cycle time, and payback.

This is how a small team with no manufacturing roots outpaced better-funded competitors.

If you want the deeper story, listen to my full conversation with
co-founder and COO Albane Dersy.

❗ It’s one of the best masterclasses I’ve recorded
on building and scaling robotics in the real world.

🎥 Full episode: https://lnkd.in/dHS5WwNa

Robots that can learn new skills without robot data. Sunday Robotics shows a different path. 

That is the idea behind ACT-1.

Instead of collecting thousands of teleoperated robot demos, they train a robot foundation model from human motion alone. 

No robot in the loop, just people wearing a glove that mirrors Memo’s hand.

Why this matters:
If you can learn from human motion at scale, you remove the biggest bottleneck in home robotics. You do not need fleets of teleop robots in real kitchens or bedrooms. You only need people showing skills with the glove.

What ACT-1 can already do:
	•	Clean a full dinner table
	•	Fold laundry for hours
	•	Assemble boxes
	•	Make espresso with long, uninterrupted runs
	•	Improve through RL (their π*0.6 model more than doubled throughput)

The interesting part is the data pipeline.

More than 500 people collected human motion demonstrations with the glove. The system builds a general understanding of objects and simple physics from that data, then transfers it to Memo.

If you want to dig into the technical write-up:
https://lnkd.in/dfWwST7M

Applications for the 2026 home beta (50 households) are open:
https://lnkd.in/dciP_gSp



——

Weekly robotics and AI insights.
Subscribe free: scalingdeep.tech

Bioinspired robot:
Fly - Roll - Walk - Crawl [Paper ⬇️]

Multi-Modal Mobility Morphobot (M4), a revolutionary robot inspired by nature's most adaptable creatures.

✅ Capable of multiple forms of movement: flying, rolling, crawling, and more.
✅ Features adaptive appendages that function as wheels, thrusters, and legs.
✅ Equipped with advanced sensors and autonomous capabilities for navigating complex terrains.

M4 shows the future of robotic locomotion, seamlessly adapting to diverse environments.

📍Paper: https://lnkd.in/ev9p9HDT

Caltech: https://lnkd.in/dgH7kKZH

Funded by 
NASA Jet Propulsion Laboratory






—-

Weekly robotics and AI insights.
Subscribe free: scalingdeep.tech

Hardware Production 🤝 Robotics

But these motors... are CRAZY synced:

CNC motor synchronization helps make sure that different motors in a machine work together smoothly so that the machine can cut and shape materials with perfect accuracy.

This is really important for making sure the machine does its job correctly.

✅ Helps the machine's parts move together perfectly, reducing mistakes and making things more accurate.

✅ Prevents the machine from getting damaged by keeping everything aligned.

✅ Allows the machine to work faster by making sure the motors are in sync.

Synchronized CNC motors make sure machines work better, last longer, and create more precise products.





—-

Weekly robotics and AI insights.
Subscribe free: scalingdeep.tech

Music tools usually live on flat screens.
This is an Apple Vision Pro app that turns music production into a modular, spatial system.

Sounds are blocks. 

Arrangements become structures. 

The composition exists as a physical object in the room, not a timeline on a screen.

OPERATOR treats music as something you build in space.

What stands out is not complexity, but restraint. Simple blocks, a snake-like sequencer, and interactions that feel closer to assembling than editing. 

No DAW metaphors. 

No menus overload.

It feels less like software and more like a spatial instrument. 

A good example of how XR can rethink creative tools instead of porting old ones into 3D.

Built by FUN AND AWE  
Saying hi to founder Maxim Zhestkov ! 




—-

Weekly robotics and AI insights. 
Subscribe free: scalingdeep.tech

taplio

The Battle

Alexander van der Steeg 🚀

Ilir Aliu

Followers

Posts

Engagements

Likes

Performance ScoreIlir Aliu's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Alexander van der Steeg 🚀

Ilir Aliu

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Ilir Aliu's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreIlir Aliu's LinkedIn Performance