Roof tile manufacturing
Catch glaze pinholes, edge chips, body cracks, shade drift, and warpage before tiles leave the kiln line.
Automated quality inspection for roof-tile production, running on a refurbished iPhone alongside your press, kiln exit, and packing line.
What is automated quality inspection for roof-tile production?
AI defect detection for roof-tile manufacturing uses a camera and a vision model to watch every tile as it leaves the press, the glaze application, the kiln exit, and the packing station, and to flag non-conforming pieces before they reach the pallet. Instead of an operator at the unloading robot or rigid rule-based vision, the model learns the clay body, glaze chemistry, profile geometry, and shade band of your SKU portfolio, and applies a consistent visual checkpoint across shifts, line speeds, and product changeovers.
Roof tiles are particularly hard to inspect at line speed because the natural variation of a clay body changes inside the same batch by design, the glaze finish reads differently across sand-faced, smooth, and engobe-coated ranges, and the hairline crack that ruins a roof installation looks identical to a normal surface texture under shed lighting. Rule-based vision built around a single profile breaks the moment you swap to a different colour, a different glaze recipe, or a different mould. AI-led inspection handles those variations because the model learns from real production frames rather than from a fixed threshold.
The result is an automated visual checkpoint that complements your end-of-line sample and gives you a tile-by-tile image record. When a roofer query comes back six weeks later, you can pull the frames from the exact production window and either confirm the defect or push back with evidence.
Defects we catch on roof-tile production lines
Body cracks and hairline fractures
Body cracks are the structural splits in the green tile or the fired piece, caused by drying-rate gradients, kiln thermal shock, or rough handling at the press unloading. Hairline fractures often run from a corner along a stress line and stay invisible until the tile flexes on a roof batten. Operators at the kiln exit catch the obvious splits but miss the early-stage hairlines that look like normal surface striations. The AI model learns the in-spec body texture and detects the local crack signature long before the line operator can. Tiles are flagged, the operator checks the press cycle and the kiln ramp, and the rejected pieces get diverted before they stack.
Edge chipping and corner breakage
Edge chipping shows up as small fractures along the leading or trailing edge, caused by die wear, stack impact at the cooler, or transfer mismatch between conveyors. Corner breaks ruin the tile for a visible eaves course. Manual operators catch the worst cases at the packing table but miss the borderline edges that pass the unloader and fail when the roofer cuts a half-tile. The AI model picks up edge texture in a single frame and flags any tile that exceeds your acceptance threshold, with the frames available so you can adjust die-change cadence or transfer alignment before the next stack ships.
Glaze pinholes and crawling
Glaze pinholes are tiny voids in the glaze film, caused by trapped gas, contaminated body surface, or undercured glaze viscosity, and crawling is the inverse where glaze beads up and exposes the body. Both create water ingress points that survive the kiln but fail the freeze-thaw warranty test. The defects are nearly invisible at the kiln exit but ruin the tile in service. The AI model holds the visual signature of an in-spec glaze for each colour and flags pinhole density and crawling clusters as soon as the local pattern deviates from spec.
Colour and shade drift
Shade drift is the gradual deviation of fired colour caused by raw-material batch variation, kiln atmosphere change, or oxide-feed drift on the glaze line. The worst cases survive the QC sample because they sit between the four corners the operator inspects, and a roof installed from a drifted pallet shows visible patches across slopes. The AI model holds a learned reference shade per range and flags drift as soon as the local colour delta exceeds your spec, giving the line a chance to correct upstream conditions before a pallet of out-of-shade tiles reaches the warehouse.
Warpage and out-of-flat
Warpage is the dimensional bow or twist that develops during drying, glazing, or kiln firing, and it shows up as tiles that rock on a batten and fail the lap test. Causes include uneven drying, kiln position effects, and clay-body composition drift. Calliper sampling at break catches the trend but misses the windows in between. The AI model picks up the surface deflection signature at the kiln exit and flags the tiles that fall outside your acceptance band before they reach the pallet, so the line can adjust drying or firing parameters early.
Cracked or missing nibs
Nibs are the integral hooks on the back of the tile that engage the roofing batten, and a cracked or missing nib means the tile cannot be installed. Causes include die wear, demoulding mistreatment, and thermal stress at firing. Manual operators check the visible face but rarely flip the tile, so nib defects often pass to the pallet. The AI model can be set up to read either the back face directly or to infer nib presence from the tile flip station, and it flags missing or cracked nibs at the unloader.
The lighting setup that makes this work on a tile line is a diffuse overhead light over the kiln exit to read body and glaze, plus a low-angle ring light at the packing station to read edge chip-out and nib geometry. An iPhone Pro with macro and wide-angle lenses handles the seven defect families from a single inspection station per critical control point. We synchronise the rig with the conveyor encoder so that flagged tiles drive a downstream divert or hold decision. We spec the optics with you during onboarding.
How Enao runs on a roof-tile line
The full hardware rig costs less than €1,000 and consists of a refurbished iPhone Pro, a diffuse overhead light with an optional low-angle ring light for edge inspection, a USB-C cable, and a mount that clamps over the kiln exit, the unloading robot, or the packing table. PLC integration is not required for the first deployment, the rig fits in a flight case, and the line keeps running while you set it up.
Onboarding is self-serve. Your line team mounts the rig, opens the Enao app, and starts collecting reference frames at the next changeover. Day one returns 80% accuracy without any prior labelling, and by day fourteen the model is operating above the manual inspector on the defect families it has seen, improving with every flagged tile that the line confirms or rejects.
Each line teaches its own model what its clay body, glaze chemistry, and tile profile look like. When you swap to a different colour or profile on the same press, the model adapts in a single shift. When you bring a sister line online with a similar product family, the second model starts from the first model's experience and the marginal effort drops sharply.
Out-of-spec tiles stop reaching the pallet, scrap is logged at the inspection point rather than at the QC office, and your operators get back the hours of attention they need for the parts of the job that still need a human, including press setup, glaze tuning, and customer complaints.
How Enao compares to manual inspection and traditional machine vision
For roof-tile producers the comparison sharpens around five dimensions.
Setup time on a roof-tile line. — Manual visual inspection: hours of training per operator, ongoing labour. Traditional machine vision (intelgic, zetamotion, Overview.ai): three to nine months of integration with a system integrator, plus a rule set per profile. Enao: deployed in a week by your own team, day one at 80% accuracy.
Hardware cost per line. — Manual visual inspection: none upfront, ongoing labour cost. Traditional machine vision: €40,000 to €200,000 per line for industrial cameras, structured lighting, and integration. Enao: under €1,000 per line with a refurbished iPhone Pro, lamp, and mount.
Handling new colours, profiles, and glazes. — Manual visual inspection: re-train operators for every new SKU. Traditional machine vision: rewrite the rule set per recipe, often outsourced to the integrator. Enao: re-teach the model on new profiles and glazes in a single shift, no code to touch.
Detection accuracy on subtle pinholes and shade drift. — Manual visual inspection: high at shift start, drops measurably after three hours. Traditional machine vision: strong on dimensional checks, weak on subtle pinhole density and shade drift. Enao: learns surface and edge signatures from reference frames and holds accuracy across shifts and runs.
Who runs it. — Manual visual inspection: trained operator at the unloader. Traditional machine vision: system integrator or a specialised vision engineer. Enao: your line team, no external specialist required.
Builder merchants and roofers change vendors over the cost of a cracked pallet, and the cost of a chargeback or a quiet specification swap sits well above the cost of an iPhone-based inspection rig. Enao is built for that gap.
