- Video analytics on existing cameras: the chain from RTSP stream to event Computer vision How to set up real-time analytics without replacing the IP cameras already in the field: stream intake, GPU frame decoding, detection and tracking, rule engine and event output. A five-link chain that has settled over five years of deployments.
- Edge or cloud? Choosing where to process in computer-vision projects Computer vision Should video be processed on site or sent to the cloud? The decision we reached from the bandwidth arithmetic, latency, data protection and three-year cost, and the exceptions to it.
- Getting people counting to 96%: field calibration notes Computer vision A model that scores 99% in the lab drops to 85% in the field. What closes the gap is not the model but the camera angle, the position of the counting line, night data and the measurement method. Notes kept on the way to 96.2% in stores.
- Data protection and video analytics: building analytics without face recognition Computer vision The five questions a legal team asks in video-analytics projects and the answers we give on the engineering side: where does the video go, what is stored, who has access, is there face recognition, how long is it kept.
- Forklift–pedestrian proximity detection: measurable safety rules in the warehouse Computer vision Measuring the distance between a forklift and a pedestrian with a camera: metres not pixels, a velocity vector not a single frame, graded warnings not an alarm. The engineering behind the rule set we built for a logistics site.
- Combining IoT and camera analytics: MQTT and an event-driven architecture IoT A camera is a sensor, the richest one. When it joins the same event bus as temperature, door contacts and PLC signals, information appears that none of them could produce alone. The event schema, MQTT topic structure and integration lessons we use in the field.
- Measuring OEE on the production line with cameras Computer vision Two of OEE's three components can be read from the PLC; the third, and the reason behind stoppages, is usually hidden in the camera. What we learned extracting machine downtime, operator presence and cycle time from video.
- Keeping detection models alive in the field: drift, retraining and dataset discipline Artificial intelligence Training the model is the shortest part of the project. The hard part is keeping a model that runs in 40 locations accurate for months while season, light, uniforms and layout change. The data and versioning discipline we settled on over five years.
- From enterprise IT management to AI entrepreneurship: what five years taught me Leadership After ten years running the IT of a retail chain, I founded my own company in 2021. What I noticed after moving from the customer's chair to the supplier's: the technology was the easy part.
Older articles (2018–2021) on HTML5, PowerShell, VBScript, SQL Server and Active Directory are available in Turkish only. Türkçe →