The same platform, configured for bin picking.
No separate system, no custom rig. You describe the job, and the same ThRiVer platform runs as a bin-picking cell: it finds each part in the pile, plans a collision-free path, and picks it, on a real robot. When the job changes, you describe the next one.
Automation should not need a new project for every job.
Traditional robot cells do one task. Programming them takes an integrator, a spec, and weeks. Change the part, the layout, or the job, and you start over. That math only works for high-volume, never-changing lines. It leaves everyone else doing by hand what a robot could do, if only setting it up were not so hard.
The robot arm was never the bottleneck. Reprogramming it is.
Effecon changes how you set a robot up.
The platform that makes it that simple.
ThRiVer is Effecon's physical-AI platform, the brain inside the workcell. You set it up by describing the job, and it handles the rest. Its name is what it does:
Understands your brief and sees the scene.
Plans safe, collision-free motion, simulated first.
Runs on your robot, cycle after cycle, and proves it.
Describe the job. The cell does the rest.
Brief ThRiVer the way you would brief a new operator. "Pick the parts from this bin and place them on the conveyor." "Stack these boxes four to a layer, three layers high." ThRiVer works out what you meant, builds the task, and shows you the plan before it moves a joint. No code. No teach-pendant programming. No waiting on an integrator.
You brief it in plain words, and it turns that into the robot's job.
One workcell. Many jobs.
Bin picking today, palletizing tomorrow, packing the day after, on the same cell. Retask ThRiVer by describing the new job instead of re-engineering the line. And because it is vendor-agnostic, the same platform runs on any standard industrial arm, not locked to one robot brand, whether we supply the cell or you run your own.
Explore applications →It sees parts it has never seen before.
ThRiVer detects and locates new parts on day one, in clutter and under changing light, with no training data and no per-part tuning. It estimates each part's position and orientation, then plans collision-free motion for a workspace that is never quite the same twice. This is the hard part of physical AI, and it is what runs when you press start.
It spots parts it has never seen, works out how they are sitting, and plans a safe path.
Performance, measured on every cycle.
Every cycle ThRiVer runs is measured: throughput, success rate, cycle time, and where time is lost, shown on a live dashboard. Return on investment becomes a figure that can be verified, not a claim.
Why Effecon.
Most systems solve one piece: a vision stack, a no-code teacher, a picking service, or a single robot brand. Effecon brings them together, on hardware you can afford, at a scale an SME can justify.
Set up by conversation
Describe the job in plain language. No code, no teach pendant, no waiting on an integrator.
One platform, many jobs
Bin picking, pick and place, palletizing, packing on one cell. Add a job by describing it, not by buying a machine.
Vendor-agnostic
Runs on any standard industrial arm, not locked to one robot brand. One we supply, or one you already run.
Handles the hard parts
New, jumbled, high-mix items with no datasets and no per-part tuning. It adapts to what is in front of it, instead of needing a fixture or a retrain for each part.
Built for SMEs
Scoped and priced for high-mix, lower-volume lines. Start with one job and one cell, not a factory rebuild.
Gets better every run
Every cycle is measured and recorded, so the platform keeps improving at exactly the jobs your line runs.
Backed by Europe's deep-tech ecosystem.
Buy the cell, or license the software.
ThRiVer Turnkey
Effecon delivers the complete cell: robot, vision, gripper, integration, and support. For teams without in-house robotics.
See turnkey →ThRiVer Software
License the AI and build your own cells, on your SKUs, scenes, and tasks. For integrators and OEMs.
See the software →Start small. Prove it. Then scale.
You do not have to automate the whole line on day one. Start with a pilot on one job, validated in simulation first and then on your hardware, with us alongside you end to end. See the ROI, then scale to more jobs on the same platform.
Start with a pilot