ABB Robotics Invests in LandingAI for Robotic Vision Enhancements

ABB Robotics has just invested in LandingAI, and it’s a move that goes beyond throwing money at a hot startup. This is about fixing one of robotics’ most persistent headaches: vision.

Robots are good at repetition. They’re bad at adaptation. Changing the lighting in a factory, swapping packaging, or moving objects in a slightly different way can cause most vision systems to fail. Fixing that usually means long retraining cycles, data annotation hell, and high-cost engineering hours. That bottleneck is exactly where LandingAI has been pushing with its platform, LandingLens™. And now ABB wants it baked into their robotics stack.

The pitch here is speed. With LandingAI’s tools, ABB claims robot vision systems can be trained and deployed up to 80% faster.

Tasks that once demanded weeks of tuning could take hours. Integrated with ABB’s RobotStudio® and digital twin software, the goal is to make robot vision less of a science project and more of a plug-and-play capability.

Why does it matter?

Because in industries like logistics, food and beverage, or healthcare, conditions change constantly. A robot that can’t adapt quickly is a liability or worse, a life hazard.

ABB hasn’t limited itself to robots that follow a basic pattern; instead, they’re selling intelligence that scales. By partnering with LandingAI, they’re trying to make adaptability a native feature rather than an afterthought.

Of course, the big claims will face real-world friction. Factories are chaotic. Edge cases always appear. But the direction is clear: robotics isn’t just about mechanical precision anymore, it’s about perception. The organization that makes vision easy to train and cheap to deploy wins.

ABB’s investment in LandingAI is a bet that vision, not motion, will decide the next decade of robotics. And they don’t want to be on the sidelines when that shift happens.

Invisible Technologies Raises $100 Million. Promises Next-Gen AI Infrastructure for Enterprises.

Invisible Technologies has just raised $100 million in growth funding, and the number isn’t the headline.

Most AI companies pitch the same gospel: plug in our model, run your data, watch the magic. Invisible doesn’t bother with that story. They’re building something far less glamorous and far more essential, the plumbing of AI. The messy, unsexy, but absolutely critical infrastructure that enterprises need if they want AI to actually work in the wild.

Think about it. Enterprises don’t run on clean, standardized data. They run on decades of legacy systems, contradictory processes, and edge cases that can kill efficiency. Drop a shiny new model into that environment, and it chokes. Invisible is trying to solve that choke point, taking AI “from A to B,” as they put it.

The funding round, led by Vanara Capital, brings Invisible’s total raise to $144 million. In parallel, their revenue doubled year-on-year to $134 million in 2024. And it says something simple: big organizations are desperate for AI that doesn’t just demo well but survives contact with reality.

Invisible brings its expansive stack to the table:

1. Neuron to wrangle chaotic data,
2. Atomic to map how businesses actually operate,
3. Synapse to keep models honest with feedback and evaluation,
4. Axon to run the agents that do the work, and
5. A global Expert Marketplace to put humans where judgment still matters.

Call it modular AI infrastructure. Call it an antidote to the “black box” trap. Either way, this isn’t hype. It’s the groundwork for AI at scale.

And that’s the real signal in this $100 million raise: the future of enterprise AI won’t be won by the flashiest demos, but by the companies building the pipes, the maps, and the muscle to make AI reliable where it counts. Invisible isn’t promising an AI that will act as the panacea for all problems.

They’re promising something rarer, an AI that actually works.