Live plant dashboard
Every machine on one screen: running, idle or down, with OEE updating every few seconds. TV-wall mode for the shop floor.
Tap any machine for details · demo data
Start with real-time OEE, downtime and quality on every machine, live in days. Then switch on AI insights and AI Vision as you grow, on the same platform and the same data. No rip-and-replace, ever.
Built for manufacturers who run on uptime
average OEE uplift in 90 days
unplanned downtime
machine-to-screen refresh
of manual reporting saved per week
The hidden factory
The gap isn't in the machines. It's in the minutes nobody sees: small stops, slow cycles, changeovers that run long, and rejects found too late.
Where a typical 8-hour shift goes
40% of planned time is lost
Illustrative · Six Big Losses
Start with your problem
Pick the question that keeps you up at night. We'll show you the part of Zelar OEE that answers it.
The platform
Every machine on one screen: running, idle or down, with OEE updating every few seconds. TV-wall mode for the shop floor.
Tap any machine for details · demo data
A × P × Q, production, energy and cycle-time trends by machine, shift or day, plus plain-language insights.
One machine, one job, one target. Operators see live count vs plan and flag issues in two taps.
Every stop is captured automatically. Operators tag the reason, so you finally know why.
Log every reject by reason in two taps. Quality % and the defect Pareto update instantly, so problems are fixed in the shift, not next morning.
A camera and a trained AI model check every part as it passes: good or defective, with the flaw boxed and logged in real time.
Watch any machine live next to its OEE data, so you can check what really happened from anywhere.
Hourly shift Excel reports and daily summaries go out automatically to the right inbox.
Plan shifts, breaks and jobs. OEE is measured against the plan, not a guess.
AI breaks losses into the Six Big Losses, flags machines at risk and ranks fixes by rupees recovered, so you always know what to do next.
Operator view
The whole screen turns green, amber or red with the machine's state. Live count against target, hourly charts, OEE and downtime, and one big button to log a rejection. No training needed.
Recreated with demo data · press “Log rejection” or switch the machine state
AI use cases
The AI runs on the same machine and camera data you collect from day one. Switch on each use case when you're ready, with no new hardware and no data migration.
AI insights · 2 available today · 4 on the roadmap
Every lost minute is sorted into the Six Big Losses and priced, so you see where the money actually goes.
Changeovers on PRS-11 cost ₹4.2 L last month, 38% of all losses.
The AI ranks improvement actions by rupees recovered, so the team works on the biggest win first.
Fix the IMM-07 feeder micro-stops first: +₹2.8 L a month.
Spots the early signs of a breakdown, such as drifting cycle times or rising micro-stops, before the machine stops.
IMM-07 is likely to stop in the next 2 hours. Cycle time is up 12%.
Ask a question in plain language and get an answer built from your own production data.
“Why was OEE low on Shift B yesterday?” → 3 die changes ran 40 min over plan.
A short written summary of every shift, ready before the handover meeting starts.
Shift A: 96% of target, 3 stops, PRS-11 needs a die check.
Suggests realistic hourly targets for each machine and part, based on your best real shifts.
Target for GR-4021 on CNC-04: 54/hr (best shift: 57/hr).
AI Vision
Your existing IP cameras become sensors. Scroll to watch the AI work on real shop-floor footage: counting, checking quality and keeping people safe.

Real frame from a live customer deployment

Real frame from a live customer deployment
An overhead camera and a trained model count finished parts and full crates every 10 seconds, all shift. No manual tally, no missed crates.
Gear inspection table · 18/36 in crate · 1,184 parts today.
11:43:09 · 16 parts ✓
11:44:29 · 17 parts ✓
11:45:49 · 18 parts ✓Real frames from a live customer deployment
Every count is saved with an evidence photo, parts outlined and number stamped. Disputes with customers or between shifts end in seconds.
11:44:29 · crate 1 · 18 parts · evidence photo saved.

Photo: General Motors · CC BY 3.0 · Wikimedia Commons
The camera checks every part and panel on the line and flags dents, scratches, chips and cracks before they reach the customer.
DEFECT · surface dent on hood · 94% · body held for rework.

Photo: Office of the President, Taiwan · CC BY 2.0 · Wikimedia Commons
Spots smoke and flames on any camera within seconds, often before a ceiling sensor reacts, and raises the alarm with the exact location.
SMOKE · Block C roof · 94% · alarm raised, safety officer called.

Photo: Sean McGrath · CC BY 2.0 · Wikimedia Commons
Checks hard hats, safety glasses and gloves at machines that need them, and alerts the supervisor the moment something is missing.
Saw station 2 · hard hat ✓ · glasses ✓ · gloves missing ✕.

Photo: CC0 · Wikimedia Commons
Confirms every component is present and in the right place before the part moves on, so a missing piece is caught at the station, not at the customer.
Board #B-118 · 41/42 components · C14 missing · flagged.
ROI calculator
Move the sliders to match your plant. We'll estimate what low OEE costs you every month, and what a 10-point improvement is worth.
Assumes 26 working days a month. “Output value per machine-hour” is what one machine produces in an hour running perfectly. This is an estimate, not a quote.
Lost to low OEE every month
₹0
9,984 planned machine-hours · running at 58% OEE
Recovered at +10 OEE points
₹0/ mo
That's per year
₹0
How it works
No PLC rewrites, no rip-and-replace. A small gateway reads what your machines already know and sends it to the platform securely.
We fit a gateway to your existing machines. No PLC rewrites, no production stop.
Live OEE, status and counts go up on every screen, from the operator's tablet to the TV wall.
Downtime reasons, the Six Big Losses and AI insights show where the time goes.
Act on the biggest losses first, then track the gain shift by shift and month by month.
Integrations
Production data flows out to your ERP, BI dashboards and team chat, so OEE becomes part of how the whole business plans, not another screen to check.
Included from day one
ERP, BI and team chat
Enterprise systems and data
Need something else? The REST API and webhooks connect to almost anything. Names shown are trademarks of their owners.
Industries
Press, machining & assembly lines
Injection & blow moulding cells
Filling, packing & batch lines
Compression, coating & packing
Converting & print presses
Forging, casting & CNC
Proof over promise
Role-based access for operators, supervisors and managers. Encrypted sessions, and your data stays yours.
Installable app that keeps working on patchy Wi-Fi, a TV-wall mode, and big, glove-friendly controls.
OEE follows the standard Availability × Performance × Quality formula and is checked against reference data.
“For the first time, my supervisors and I argue about solutions instead of about whose numbers are right.”
“The downtime Pareto showed us our changeovers were the real problem, not the machines.”
“Shift reports used to take all morning. Now they're in my inbox before the shift meeting.”
Overall Equipment Effectiveness is the share of planned production time that is truly productive. It multiplies Availability (was it running?), Performance (was it running at full speed?) and Quality (were the parts good?). 100% means only good parts, as fast as possible, with no stops.
Book a 30-minute walkthrough. We'll show you the platform running on live machine data and map out a pilot for your plant.