Shim hours per join
34h
▼ -64% decrease
Three deployments, measured against signed baselines
Each of these started as a line assessment on one station. Each has a baseline that was agreed in writing before a single agent ran, measured from the same telemetry the agents use.
| Metric | Magnitude of change (%) |
|---|---|
| Shim hours per join | 64.0% |
| Delivered rate | 38.0% |
| Escapes to next station | 71.0% |
| Join cycle time | 47.0% |
| Takt adherence gain | 11.4% |
| FOD events | 84.0% |
Airframes assembled on Rivetira across commercial, defense, business-jet, eVTOL and space-structure programs: 8 factories, 70 assembly stations.
Meridian Aerostructures
Calder Aviation Group
Northvane Defense
Solstice Jets
Helion Air Mobility
Orbital Frame Systems
Removing hand-fit shimming from a narrowbody wing box
Meridian Aerostructures · Tier-1 aerostructures. Predictive shimming cut fit-up from 3.4 iterations to a single machined pass across 14 stations.
Gap prediction MAE
0.031 mm
▼ -58% decrease
Join cycle time
2.6 d
▼ -47% decrease
Right-first-time
98.7%
▲ +8.4% increase
The problem
A narrowbody wing box join was taking 3.4 fit-up iterations. Each iteration meant separating the structure, hand-measuring the gap with feeler gauges, machining or stacking shims, and re-presenting. Fourteen stations, 94 shim hours per join, and a fit that still varied by mechanic.
What we deployed
Metrology & Predictive Shimming ran in shadow for nine weeks across 1,900 joins, learning the relationship between as-built part geometry, fixture state, thermal condition and the resulting gap field. Once gap prediction MAE fell below 0.04 mm it was promoted to advisory, then to closed loop driving the 5-axis shim machining cell directly.
The result
Fit-up became a single machined pass. Shim hours per join fell from 94 to 34, join cycle time from 4.9 days to 2.6, and right-first-time on structural joins reached 98.7%. The wing box stopped being the constraint on the line.
| Week since kickoff | Gap prediction MAE | Hand-measurement error |
|---|---|---|
| Wk 1 | 0.112 mm | 0.074 mm |
| Wk 5 | 0.089 mm | 0.076 mm |
| Wk 9 | 0.068 mm | 0.073 mm |
| Wk 13 | 0.052 mm | 0.075 mm |
| Wk 17 | 0.043 mm | 0.072 mm |
| Wk 21 | 0.037 mm | 0.074 mm |
| Wk 25 | 0.033 mm | 0.073 mm |
| Wk 29 | 0.031 mm | 0.074 mm |
Meridian Aerostructures: the underlying numbers
| Measure | Value (mixed units — see table) |
|---|---|
| Fit-up iterations (before) | 3.4 |
| Fit-up iterations (after) | 1.1 |
| Shim hours (before) | 94.0 |
| Shim hours (after) | 34.0 |
| Join cycle days (before) | 4.9 |
| Join cycle days (after) | 2.6 |
“We stopped arguing about whether the gap was 0.4 or 0.6 millimetres. The twin predicted it, the shim came off the machine right, and the join closed in one pass.”
−64% shim hours per join
- Segment
- Tier-1 aerostructures
- Stations deployed
- 14
- Program
- Narrowbody wing box
- Baseline window
- 90 days pre-deployment
- Autonomy level reached
- Closed loop
Holding takt through a single-aisle rate ramp
Calder Aviation Group · Commercial OEM. Line-and-takt balancing across 22 stations absorbed a 38% rate increase without adding positions.
Delivered rate
41/mo
▲ +38% increase
Takt adherence
96.8%
▲ +11% increase
Travelled work
−44%
▼ -44% decrease
Overtime hours
−29%
▼ -29% decrease
The problem
A single-aisle final assembly line was committed to a 38% rate increase. The plan on the table was four additional positions and a building extension. Takt adherence was 85.4% and travelled work was rising every month as jobs slipped downstream.
What we deployed
Line & Takt Optimisation ingested station cycle data from the MES across 22 stations and rebalanced work content every shift against actual, not planned, cycle times. Where a station drifted, work was re-sequenced ahead of the drift rather than after it. Robotic handling supervision removed the crane-wait time that had been absorbing float.
The result
Delivered rate went from 30 to 41 aircraft per month inside the existing footprint. Takt adherence reached 96.8%, travelled work fell 44%, and overtime for assembly mechanics fell 29%. The building extension was cancelled.
| Month | Delivered rate | Planned rate |
|---|---|---|
| Jan | 30 /mo | 30 /mo |
| Feb | 30 /mo | 31 /mo |
| Mar | 31 /mo | 32 /mo |
| Apr | 33 /mo | 33 /mo |
| May | 34 /mo | 35 /mo |
| Jun | 36 /mo | 36 /mo |
| Jul | 37 /mo | 38 /mo |
| Aug | 39 /mo | 39 /mo |
| Sep | 40 /mo | 40 /mo |
| Oct | 41 /mo | 41 /mo |
Calder Aviation Group: the underlying numbers
| Period | Value-added time | Wait for parts / crane | Rework & travelled work | Inspection wait | Total |
|---|---|---|---|---|---|
| Pre-deployment | 58% | 21% | 14% | 7% | 100% |
| Month 3 | 63% | 17% | 12% | 8% | 100% |
| Month 6 | 69% | 13% | 10% | 8% | 100% |
| Month 9 | 74% | 10% | 8% | 8% | 100% |
| Month 12 | 78% | 8% | 8% | 6% | 100% |
“The rate ramp was going to cost us four more positions. Instead the line rebalanced itself every shift and we found the capacity inside the stations we already had.”
+38% delivered rate
- Segment
- Commercial OEM
- Stations deployed
- 22
- Program
- Single-aisle final assembly
- Baseline window
- 90 days pre-deployment
- Autonomy level reached
- Closed loop
Driving airworthiness escapes toward zero on a tactical airlifter
Northvane Defense · Defense prime. Hole-and-fastener inspection caught 412 conditions before the structure left the station.
Escapes to next station
−71%
▼ -71% decrease
FOD events
2/mo
▼ -84% decrease
NCR cycle time
−52%
▼ -52% decrease
Inspection coverage
100%
▲ +71% increase
The problem
A tactical airlifter program inspected 29% of holes by sample. Escapes were reaching the next station, and each one meant disassembly, disposition and a nonconformance record. FOD events were running at twelve a month. The program office wanted an audit trail nobody could produce in under two weeks.
What we deployed
Hole & Fastener Inspection deployed air-gapped across nine stations. Every drilled hole was measured for diameter, countersink depth, perpendicularity, burr and surface condition; every fastener was checked for seating, flushness and grip length. FOD sensing ran continuously on the same imagery. Nothing left the enclave.
The result
Inspection coverage went from 29% sampled to 100%. Escapes to the next station fell 71%, FOD events from twelve a month to two, and NCR cycle time from raise to disposition fell 52%. A complete airframe trace is now produced in ninety seconds.
| Quarter since go-live | Escapes to next station | Conditions caught in station |
|---|---|---|
| Q1 | 84 | 31 |
| Q2 | 71 | 68 |
| Q3 | 52 | 94 |
| Q4 | 41 | 112 |
| Q5 | 31 | 106 |
| Q6 | 24 | 98 |
Northvane Defense: the underlying numbers
| Segment | Share |
|---|---|
| Countersink depth | 118 |
| Hole diameter | 94 |
| Perpendicularity | 71 |
| Burr / surface condition | 58 |
| Fastener seating | 41 |
| FOD in structure | 30 |
“Every hole is now inspected, not one in four. My airworthiness record writes itself, and I can hand an auditor a complete trace in ninety seconds.”
100% inspection coverage
- Segment
- Defense prime
- Stations deployed
- 9
- Program
- Tactical airlifter
- Baseline window
- 90 days pre-deployment
- Autonomy level reached
- Supervised (air-gapped)
What the three deployments have in common
A signed baseline first
Every deployment locked its baseline in writing before a single agent ran. Without that, a result is an anecdote.
Shadow mode before advisory
Agents predicted for weeks while humans worked normally. Promotion only followed measured accuracy against physical reality.
One station, then the line
None of these started as a factory rollout. Each proved a single cell first, then expanded on the strength of the number.
Where the money actually comes back
Rivetira is priced on the metric it moves. This is the value stack from a representative 14-station wing-box deployment.
| Value source | Annualised value (USD) |
|---|---|
| Shim labour removed | $4.9M |
| Rework hours avoided | $3.6M |
| Rate capacity unlocked | $3.1M |
| Scrap & escapes avoided | $2.2M |
| Overtime reduction | $1.4M |
| Inspection labour | $0.9M |
Annualised value
$16.1M
Platform cost
$2.7M
Payback period
2.4 mo
Net revenue retention
136%
▲ +36% increase
Shadow first. Autonomy is earned, not switched on.
Every agent starts by watching. It is promoted only when its measured accuracy clears the mechanic-plus-metrology baseline and the twin agrees.
-
Weeks 1–4
Shadow
Agent observes the station, predicts every outcome, actuates nothing. Accuracy measured against what the mechanics actually do.
-
Weeks 4–10
Advisory
Agent recommends feed, shim geometry and sequence. A human accepts or rejects; every rejection becomes training data.
-
Weeks 10–20
Supervised
Agent actuates with a human in the loop and a live fail-safe stop. Airworthiness-critical dispositions still require sign-off.
-
Week 20+
Closed loop
Agent runs the step. Humans handle exceptions and the twin gates any change to the control policy.
| Week | Agent accuracy | Mechanic + metrology baseline |
|---|---|---|
| Wk 2 | 88.2% | 94.1% |
| Wk 4 | 91.4% | 94.0% |
| Wk 6 | 93.6% | 94.2% |
| Wk 8 | 95.2% | 94.1% |
| Wk 10 | 96.4% | 94.3% |
| Wk 12 | 97.1% | 94.2% |
| Wk 16 | 97.8% | 94.1% |
| Wk 20 | 98.3% | 94.2% |
| Wk 24 | 98.6% | 94.3% |
Assembly autonomy, measured on the line
Every figure below is produced by the same telemetry the agents act on — station cycle, hole quality, gap field, fastener state and conformance. Pilot and design-partner aggregate, trailing 12 months.
Holes drilled under agent control
41.6M
▲ +32% increase
Right-first-time, structural joins
97.9%
▲ +9.1% increase
Shim hours removed per join
68%
▼ -68% decrease
Assembly-line uptime
99.94%
▲ +0.3% increase
Aggregate across design-partner lines. Baselines are the same stations before Rivetira, measured over an equivalent period.
Questions engineers actually ask
Anything else goes to the people who build it. Ask an assembly engineer or read the full FAQ.
The deployments, structures and results are representative of Rivetira engagements; company names are anonymised at customer request as is standard on aerospace programs. Named references are available under NDA after a line assessment.
Because the constraints differ. A commercial ramp is throughput-constrained, a defense program is escape-constrained, an eVTOL program is industrialisation-constrained. The line assessment identifies which one you are before anyone quotes a number.
Shadow mode produces measurable prediction accuracy inside six weeks. Operational results follow the first closed-loop promotion, which has ranged from four to eleven months depending on structure complexity and accreditation requirements.
Build your own case study
A line assessment maps one station, quantifies the rework, shim and rate opportunity, and returns a modelled ROI in three weeks. No production disruption.