The robot now recognizes each panel, calculates its exact position, and corrects its own motion in real time. Result: from bottleneck to top performer within Stellantis in under 2 years. "We watched this go from the bottleneck of the plant, to, 'Boom!' We are back in business." — Sean Woodall, SVP Assembly Operations

Stellantis was told the fix was new robots with embedded vision. Hundreds of thousands of dollars each, in a plant running close to 1,000 robots. They chose differently: our camera + software on the same 15-year-old arm.

The problem: whenever a panel sat slightly out of position on its rack, the robot got confused and stopped. Sometimes it crashed for an hour. On an auto line, a bottleneck failure can cost $100,000 a minute.

Yaskawa robots now run Inbolt natively, including on moving lines. A robot-mounted camera locates the part and corrects the robot's motion in real time. That's six major brands with native support: FANUC, ABB, KUKA, UR, Comau, and now @Yaskawa_Motoman. Video below. Talk to our team to give your Yaskawa robots vision: inbolt.com/contact/ #Yaskawa #Robotics #Inbolt

The fastest-growing "bin picking" application isn't bin picking. Walk a modern auto or metal-forming plant. Half of what's labelled "bin pick" is dunnage, racks, carriers, single-layer totes, deracking from fixtures, picking off moving conveyors. The deep-bin random-orientation pick (the canonical case) is one slice of the category. That matters because vision systems get specced for the headline case. Buyers say "bin picking," vendors hear the random-bin showcase video, and the system arrives over-engineered for chaos that isn't there, or under-engineered for the variability that is. It's a spectrum. Structured at one end, fully random at the other. Most factory work sits in the messy middle, and the volume coming online from mixed-reference lines, model proliferation, and reshored brownfield expansions is growing fastest in that middle. A vision system needs to handle the full range. The question for plant engineers isn't "do we need bin picking?", it's "what's the actual structure of the picking we already call bin picking?" How does your line break down: structured, semi-structured, random? #Manufacturing #IndustrialAutomation #MachineVision #Inbolt #BinPicking #Robotics

Inbolt was named as one of 21 European startups on the rise, in the @TechCrunch article "Beyond Lovable and Mistral: 21 European startups to watch". Check it out here: techcrunch.com/2026/05/02/be… #Robotics #Automation #Manufacturing #EuropeanTech #TechCrunch #Inbolt

Do your robots fail because a part is sitting just a few millimeters off? Do moving-line tasks still rely on operators to keep production running? Are shift workers stuck handling heavy, cluttered bins? That’s where Inbolt comes in. Inbolt gives robots 3D vision so they can locate parts in real time, adapt trajectories, and execute reliably across fixed cells, moving lines, and unstructured environments. 👷Plant engineers: what’s the “simple” station in your plant that somehow causes the most chaos? #Robotics #Automation #IndustrialAutomation #3DVision #Inbolt

Ultra-fast processing is the difference between automation that works, and automation that stalls. At @Teksid, molds hit the station fast, stop unpredictably, and vary across references. The task: Level ferro-alloy. Insert filters. Hold cycle time. Inbolt’s real-time 3D vision lets a @Universal_Robot cobot detect each mold instantly and servo-correct its trajectory on the fly. • 6DoF pose in milliseconds • Continuous path correction • No indexing • No custom mechanics Fast-moving line. 9-month ROI. New mold variants programmed in <10 minutes. Thanks to @Drim_Robotics for the integration. Where is processing speed silently limiting your automation? #Inbolt #PhysicalAI #3DVision #Robotics #FactoryAutomation

More automotive OEMs are rethinking automation: bringing it back in-house to cut costs and move faster. At a @Stellantis plant in the US, a manual de-racking station handling mixed parts was automated without redesigning the line. The internal team built the cell. Inbolt added real-time 3D vision with a robot-mounted camera, allowing the robot to localize each part and adapt its motion live. Results in production: • 100% picking accuracy • Deployed in 2 days • Lower automation cost vs turnkey solutions • Faster ROI with full ownership A clear example of how in-house engineering + real-time AI vision enables faster, capital-efficient automation. Read the full case 👉 lnkd.in/eqrCaMBV Is this becoming a broader trend in automotive? #IndustrialAutomation #Manufacturing

Production lines don’t stop moving. Robots shouldn’t either. Traditional vision breaks down once parts are in motion, forcing manual work, higher costs, and quality drift. Inbolt brings real-time AI vision on the robot arm, keeping robots perfectly synchronized with moving parts: 🔹 <80 ms detection 🔹 ±0.5 mm precision 🔹 Full line speed, no stops Fewer operators. Higher throughput. Real automation built for real factories. If your line never stops, why should your robots struggle to keep up? #Manufacturing #IndustrialAutomation #Robotics #MachineVision #SmartFactory #Industry40

🚀 We’re launching Inbolt’s intelligent bin picking solution, built for the real chaos of factory floors. Traditional systems break the moment parts move, overlap, tilt, or lighting shifts. Inbolt doesn’t. 🦾🔥 Our AI vision guidance runs on the robot arm, so robots can: ⚡ See + adapt in real time ⚡ Identify any pickable side in fully unstructured bins ⚡ Re-analyze after grasp + auto-correct trajectory ⚡ Hit <1s pick cycles with up to 95% success in production All with one on-arm 3D camera. No fixed rigs. No heavy calibration. Smarter, faster, and finally robust bin picking — ready for real-world manufacturing. 👉 Full announcement + datasheet: inbolt.com/resources/newsroo… #BinPicking #3DVision #Robotics #IndustrialAutomation #SmartManufacturing #AI #Inbolt

Stop letting imperfect pallets break your automation! A Hungarian @Stellantis plant partnered with us to achieve more reliable cylinder head depalletizing, even with pallet variance. Inbolt's AI 3D Vision-Guidance gave the FANUC robot human-like vision, adapting to real-world part shifts and uneven pallets. ➡️ Higher Success Rate ➡️ More Uptime ➡️ Lower Costs Adaptable, AI grasping is the future. Read the case study: lnkd.in/efK7-BeR #Inbolt #Automation #3DVision #AI #Manufacturing

#iREX2025 ALERT: INBOLT WILL BE LIVE IN TOKYO! 🇯🇵 We'll be at the world's largest robotics show with two major demos: - Partnering with Universal Robots (Booth E7-20). - A demo with a major global robot partner (TBA!). See production-proven AI & 3D Vision guide robots to perform real-time gluing on moving workpieces. No more static automation. If you are at #iREX2025, accelerate your automation roadmap. Book a meeting with our APAC Sales Head, Takaki Sakaguchi: meetings.hubspot.com/takaki-… See you in Tokyo! #Robotics #VisionGuidance #Manufacturing #iREX2025 #Inbolt

One year ago: 15 engineers in Paris. Today: 40+ people across Paris, Detroit & Tokyo, and our 3D vision is running in some of the world’s toughest factories. Our €15M Series A didn’t just fund us. It accelerated everything. The next era of automation won’t come from adding more robots. It’ll come from giving robots vision & intelligence. That’s the future we’re building at Inbolt. #AI #Robotics #Automation #Inbolt