Industries · 6 min read
Perimeter security with fixed cameras, object detection and a drone sent to look
A frame every two seconds is enough to catch a person at the fence, a CPU is enough to run it, and the drone is the second look rather than the detector.
Andreas Ohrvall · Sep 30, 2026
Industries · 6 min read
Custom vision models for industrial robots, trained on your own castings
The plant's own castings labeled from the cell's camera, defect, orientation and grasp point as the three questions, and a new variant handled by labeling it.
Andreas Ohrvall · Sep 30, 2026
Edge · 6 min read
Edge AI fleet management for fifty runners across five plants
Fifty small boxes beside fifty recorders, one model version, and the IT patch that changes what every camera sends. What a fleet needs before the second plant.
Andreas Ohrvall · Sep 30, 2026
Edge · 6 min read
What computer vision model deployment means after the first week
Deploying a vision model is a decision about where it runs, how a new version replaces the old, and what tells you it has started to be wrong.
Andreas Ohrvall · Sep 30, 2026
Industries · 6 min read
AI in robotics after the robot ships, what the warehouse cameras keep learning
The forward camera boxed pallets and people well at the pilot site. Then the racking moved, and the edge cases the planner never saw came back for review.
Andreas Ohrvall · Sep 29, 2026
Computer vision · 6 min read
What the big vision models are for, and what runs on the camera
The big models find the frames, draw the first boxes and tighten them, and a person checks. The small model trained on those labels runs beside the recorder.
Andreas Ohrvall · Sep 29, 2026
Industries · 6 min read
Manufacturing ERP systems and the gap between the ledger and the floor
The WIP module knows what somebody typed at shift end. A camera counting units off the fixture as they happen posts the record by webhook, minutes behind.
Andreas Ohrvall · Sep 29, 2026
Industries · 7 min read
Sensor fusion with computer vision, the tag says which pallet and the camera says what happened to it
The RFID portal reads a tag at the dock door. The camera above it sees the pallet lifted, filled or left behind, and catches the tag that fell off on the way.
Andreas Ohrvall · Sep 29, 2026
Computer vision · 6 min read
What YOLO does in one pass and what that costs
One look at the frame is why a single-pass detector fits the box beside the recorder, and why it misses the small cluttered things a two-stage model finds.
Andreas Ohrvall · Sep 29, 2026
Operations · 6 min read
Computer vision on the industrial HMI, the camera's verdict on the operator screen without the noise
The verdict as a state, the frame one click away, an alarm the operator acknowledges like any other, and everything else kept off the screen.
Andreas Ohrvall · Sep 28, 2026
Labeling · 6 min read
YOLO knowledge distillation, let the big model label and the small model run
A large model boxes the warehouse frames once, a floor lead corrects them, and a compact detector trained on that set runs beside the recorder.
Andreas Ohrvall · Sep 28, 2026
Edge · 8 min read
AI cameras vs IP cameras, and what changes when the model moves to the edge
The dock camera has streamed to a recorder for six years. A model can watch that stream in the cloud, on your servers or beside the recorder. No new camera.
Andreas Ohrvall · Sep 27, 2026
Edge · 7 min read
Cloud vs on-device inference for computer vision, and why the answer is usually both
A remote substation on a thin link and a plant with a footage policy. The runner decides on site, only doubted frames leave, the cloud trains the next version.
Andreas Ohrvall · Sep 27, 2026
Edge · 8 min read
Running computer vision on the IP cameras a site already has
A store, a plant and a yard already stream RTSP to a recorder. That stream is the only requirement; the camera quote most projects begin with was never needed.
Andreas Ohrvall · Sep 27, 2026
Edge · 6 min read
Deploying computer vision models offline, where the network does not reach
A vessel at sea for three weeks and a plant with no outbound video: the runner beside the recorder, alerts on site, doubted frames leaving only on a link.
Andreas Ohrvall · Sep 27, 2026
Edge · 7 min read
On-premise computer vision inference when the footage cannot leave the fence
A utility, a plant and a vessel whose security teams refuse outbound video. The model runs on your servers or beside the recorder; only doubted frames leave.
Andreas Ohrvall · Sep 27, 2026
Operations · 6 min read
PLC and computer vision integration, where the model fits next to the controller
How a scan cycle works, why the PLC keeps the reject arm, the detection arriving as a webhook into the broker the PLC reads, and the latency a line can absorb.
Andreas Ohrvall · Sep 27, 2026
Operations · 6 min read
Video analytics with vision AI, the three stages every camera system shares
From a motion rule that fires at dusk to a model that answers whether the package is damaged, and where the runner and the review queue sit in it.
Andreas Ohrvall · Sep 27, 2026
Industries · 6 min read
Pipeline inspection with camera nodes that watch for the crack before it is a leak
Cameras on the exposed spans, classes written with the integrity team, the alert raised at the node with the frame, and rain showing in the corrections first.
Andreas Ohrvall · Sep 26, 2026
Industries · 6 min read
Warehouse rack occupancy detection from a camera on the forklift, so an empty bay is a finding and never a guess
Rack slots and pallets as separate classes, occupancy written back per position with the frame, and stock staged in the aisle as the first failure to rule out.
Andreas Ohrvall · Sep 26, 2026
Industries · 7 min read
Computer vision for autonomous mobile robots that pull pallets from dark trailers
A robot's forward cameras in a dark trailer find a worn pallet and its pose, keep people in view, and the second warehouse is a new site with its own frames.
Andreas Ohrvall · Sep 25, 2026
Industries · 8 min read
Georeferencing drone detections from a box in the frame to a GPS point the crew can drive to
Heading, altitude and field of view turn a boxed insulator into a coordinate. The log and the video disagree on time, and a re-planned flight moves every box.
Andreas Ohrvall · Sep 25, 2026
Industries · 6 min read
Pallet scanning with a camera portal that reads the label from whichever side it landed on
Cameras on four sides of the dock door find the label wherever the wrapper left it, decode the barcode, read the text, and log every scan with its frame.
Andreas Ohrvall · Sep 25, 2026
Industries · 7 min read
Automated pallet accounting from the camera over the staging zone
A polygon on the frame, every pallet tracked so it is counted once, entries and exits as the ledger, and a wash-down that nudges the camera as the failure.
Andreas Ohrvall · Sep 24, 2026
Industries · 7 min read
Computer vision for robotics, what the cameras have to answer before the arm moves
Detection, segmentation, pose estimation and depth, in that order, with the label each one needs, and a firmware update as the first failure to check.
Andreas Ohrvall · Sep 24, 2026
Industries · 7 min read
Lights-out manufacturing with a camera on the chip pile and the part seat
An unattended machining cell from 10 pm to 6 am, three cameras where an operator's eyes used to be, and the frame that reaches the on-call phone in time.
Andreas Ohrvall · Sep 24, 2026
Operations · 7 min read
Integrating machine vision with PLC, SCADA, MES and ERP systems
A cracked weld is a bit for the PLC, a thumbnail for the HMI, a record for the MES and a yield figure for the ERP. Shape it once, beside the recorder.
Andreas Ohrvall · Sep 23, 2026
Operations · 7 min read
Getting a defect detection to the PLC over MQTT
One webhook from the alert into the plant's broker, a topic per line and camera, and the PLC and the SCADA screen subscribe like any other client.
Andreas Ohrvall · Sep 23, 2026
Industries · 6 min read
Production line monitoring with the cameras already over the line
The cameras were installed for disputes. A model turns a missing cap, a low fill and a count into events for the line lead. The mount decides whether it holds.
Andreas Ohrvall · Sep 23, 2026
Operations · 6 min read
Deploying a computer vision model to a second plant, beyond the weights file
What the second plant needs beyond a weights file: recorded thresholds and the test set behind them, the alert template, acceptance criteria, its own frames.
Andreas Ohrvall · Sep 23, 2026