Control carbody welding and millimetre geometry before shell defects reach final assembly.
Carbra is the carbody fabrication agent.
Products
Buy the step that hurts. Add the rest when it pays.
Railira ships as five products mapped to the plant, not to abstract capabilities. Carbra controls carbody welding and geometry, Bogira controls bogie and running-gear assembly, Coatyx controls livery and protective coatings, Fitura manages fit-out, marriage and the as-built train twin, and Taktos sequences the plant on takt. Start with one on the Process tier and expand as each proves its number.
Illustrative scenario: Car 07 of a 4-car metro consist · shell line B · bogie shop · booth 2 · marriage station, one car, one approval gate
The five
One stack, five products, one record
Carbra controls the shell: EN 15085 welding and millimetre geometry, the structure everything else fits to. Bogira assembles the running gear to torque, press-fit and ride-quality spec. Coatyx applies livery and protective coatings right the first pass. Fitura aligns carbody, bogie, interiors and systems and hosts the as-built train twin. Taktos sequences production, kits, AMRs, commissioning and tests on takt. Data flows Carbra to Bogira to Fitura to Taktos; Coatyx cure windows constrain Taktos; the twin lives in Fitura and is fed by all.
Assemble bogies and running gear to torque, press-fit and ride-quality spec.
Bogira is the bogie-and-running-gear assembly agent.
Apply rail livery and protective coatings right the first pass.
Coatyx is the coat-and-livery agent for rolling-stock plants.
Align carbody, bogie, interiors, systems and the as-built train twin.
Fitura is the fit-out, marriage and train-twin product.
Sequence railcar production, JIS feeding, AMRs, commissioning and dynamic tests on takt.
Taktos is the line-and-logistics agent.
Carbra · product family colour #178A47
Carbra
Control carbody welding and millimetre geometry before shell defects reach final assembly.
The problem
Rail carbodies are huge welded structures, not small parts. Steel shells and aluminium extrusion bodies must meet EN 15085 weld quality, EN 12663 fatigue expectations, EN 15227 crashworthiness intent, door aperture tolerances, roof, sidewall and underframe geometry, and a livery-ready surface. A heat-input drift, fixture shift, missed seam, extrusion mismatch or framing build-up can move a shell out by millimetres and force days of rework later in fit-out or commissioning.
What Carbra does
Carbra is the carbody fabrication agent. It fuses robotic weld parameters, seam vision, thermal signatures, fixture data, laser trackers, point clouds and non-conformance history to predict weld defects and geometry drift in real time. It recommends weld sequence, clamping, rework priority and inspection coverage inside approved control-plan envelopes, while GPU FEA surrogates estimate EN 12663 fatigue and EN 15227 crash-load sensitivities before the shift.
Who it is for
Rolling-stock OEM carbody shops building locomotives, high-speed and intercity trainsets, metro and commuter cars, passenger coaches and freight wagons; contract carbody fabricators; welding-cell integrators and rail manufacturing-engineering teams.
Key features
- EN 15085 weld-defect vision for seams, plugs, brackets and extrusion joints
- Millimetre carbody alignment prediction from laser trackers, point clouds, fixtures and weld history
- Weld-sequence and heat-input recommendations with human approval gates
- H100 or HGX FEA surrogates for carbody fatigue and crashworthiness sensitivity
- Feeds shell geometry and weld genealogy into Fitura, Taktos, Coatyx and the conformance record
NVIDIA stack
Named in the product brief as the design
Jetson Orin / ThorH100 / HGXDeepStream / MetropolisHoloscanTensorRTTritonNVIDIA AI EnterpriseNIM / NeMo
Planned
DGX for cross-program weld and geometry fine-tuning · plannedSupervised weld-sequence optimisation inside approved control plans · planned
Why it needs a GPU
Real-time weld vision plus large-structure point clouds and FEA surrogates make GPUs mandatory at the edge and at training scale. CPU-only cannot hold the seam-alert latency or train carbody-scale surrogates.
MVP scope and first metric
Shadow mode on one instrumented shell line: prove weld-defect and geometry-drift prediction against the shop's own rework and acceptance records, then graduate to supervised recommendations and bounded control inside approved rail work instructions. First metric: weld rework and geometry escapes on a named line.
Tier. Process tier, $16,000 per carbody line or area per month, expanding to Plant and Enterprise.
Open questions
- Per-plant sensor coverage, OT access and acceptance-data availability vary
- Cross-plant model transfer waits on customer IP and data agreements
Bogira · product family colour #178A47
Bogira
Assemble bogies and running gear to torque, press-fit and ride-quality spec.
The problem
A bogie is the rail vehicle's safety-critical running gear. Wheelset press-fit, axle boxes, primary and secondary suspension, brake rigging, traction motors, gearboxes, yaw dampers and torque-to-spec joints all affect ride quality, reliability and acceptance. Assembly drift may only appear during brake tests, dynamic ride-quality runs or warranty service.
What Bogira does
Bogira is the bogie-and-running-gear assembly agent. It ingests press-fit curves, torque traces, bearing data, motor and gearbox mounting telemetry, brake-system checks, machine vision and inspection records to detect assembly anomalies, predict ride and brake risk, and guide technicians through torque, shim, suspension and mounting decisions inside approved work instructions.
Who it is for
Rolling-stock OEM bogie shops, locomotive and metro truck builders, overhaul facilities, and manufacturing-engineering teams responsible for running gear, brakes, wheelsets, suspension and traction mounting.
Key features
- Wheelset press-fit curve anomaly detection
- Torque-trace analytics for safety-critical bogie joints
- Suspension, brake, traction-motor and gearbox mounting verification
- Vision inspection for missing parts, orientation, labels and harness routing on the bogie
- Feeds ride-quality, brake and running-gear risk into Fitura and Taktos
NVIDIA stack
Named in the product brief as the design
Jetson Orin / ThorH100 / HGXHoloscanDeepStream / MetropolisTensorRTTritonNVIDIA AI EnterpriseNIM / NeMo
Planned
DGX for multi-plant running-gear model training · plannedRide-quality risk prediction from build data · planned
Why it needs a GPU
Multi-stream station vision and anomaly detection on high-volume torque and press-fit traces need edge GPUs; network learning across bogie corpora uses H100 or HGX.
MVP scope and first metric
Shadow mode on one instrumented bogie line: prove press-fit and torque anomaly detection against brake-test and inspection outcomes, then graduate to supervised guidance inside approved work instructions. First metric: press-fit and torque escapes caught before brake test.
Tier. Process tier, $16,000 per bogie line or area per month, expanding to Plant and Enterprise.
Open questions
- Per-plant sensor coverage, OT access and acceptance-data availability vary
- Cross-plant model transfer waits on customer IP and data agreements
Coatyx · product family colour #178A47
Coatyx
Apply rail livery and protective coatings right the first pass.
The problem
Rail exteriors must survive weather, ballast impact, washing chemicals, graffiti, corrosion, UV and customer livery standards. Surface prep, primer, film build, masking, cure windows and environmental conditions all interact. A coating defect can block delivery, force a repaint, delay commissioning or create lifecycle corrosion risk.
What Coatyx does
Coatyx is the coat-and-livery agent for rolling-stock plants. It fuses booth cameras, surface-prep data, thickness gauges, cure and flash telemetry, humidity and temperature, recipe history and visual inspection to predict runs, dirt, adhesion risk, colour or livery mismatch, corrosion-protection gaps and cure-window conflicts before they reach acceptance.
Who it is for
Rolling-stock OEM coating and livery shops, passenger coach, metro and transit railcar coating lines, freight-wagon coating lines, refurbishment facilities and livery suppliers.
Key features
- Surface-prep, primer, livery, film-build and cure-window monitoring
- Vision detection for dirt, runs, masking errors, colour breaks and livery alignment
- Anti-graffiti and anti-corrosion coating conformance tracking
- Flash and cure window constraints handed to Taktos
- Coating genealogy tied to IRIS / ISO/TS 22163 and customer acceptance records
NVIDIA stack
Named in the product brief as the design
Jetson OrinDGX / HGXDeepStream / MetropolisTensorRTTritonHoloscanNVIDIA AI Enterprise
Planned
Cosmos and Omniverse Replicator for rare coating-defect synthesis · plannedOVX for coating-cell and airflow simulation in the plant twin · planned
Why it needs a GPU
High-resolution coating vision and rare-defect synthetic generation require GPU edge inference and GPU training. CPU-only cannot inspect a booth at line speed.
MVP scope and first metric
Shadow mode on one coating area: prove defect and cure-window prediction against the shop's repaint and acceptance records, then graduate to supervised recipe and routing recommendations. First metric: final-livery defects and repaint events on a named booth.
Tier. Process tier, $16,000 per coating area per month, expanding to Plant and Enterprise with outcome components on repaint and delivery delay.
Open questions
- Per-plant sensor coverage, OT access and acceptance-data availability vary
- Synthetic rare-defect fidelity must be validated against real distributions
Fitura · product family colour #178A47
Fitura
Align carbody, bogie, interiors, systems and the as-built train twin.
The problem
Final rolling-stock assembly couples enormous subsystems: carbody, bogies, couplers, gangways, doors, glazing, floors, seats, cab equipment, HVAC, pantograph or third-rail gear, batteries, wiring harnesses and accessibility interiors. Misalignment or late discovery creates rework, missed acceptance windows and delivery delay.
What Fitura does
Fitura is the fit-out, marriage and train-twin product. It hosts Railira's Omniverse as-built train-and-plant digital twin, consuming Carbra shell geometry, Bogira running-gear data, Coatyx coating state and underframe and propulsion telemetry. Fitura predicts fit, wiring, accessibility, door, glazing, interior, cab and carbody-to-bogie marriage risk before work starts and guides supervised corrections.
Who it is for
Rolling-stock final assembly plants, trainset integrators, metro and transit car builders, passenger coach plants, locomotive assembly teams and commissioning engineering groups.
Key features
- Omniverse as-built train-and-plant digital twin
- Carbody-to-bogie marriage, coupler and gangway alignment guidance
- Interior fit-out validation for seats, panels, flooring, doors, glazing, cab and ADA/PRM requirements
- Underframe propulsion, HVAC, pantograph or third-rail, battery, wiring routing and continuity risk prediction
- Feeds readiness, rework and acceptance constraints into Taktos
NVIDIA stack
Named in the product brief as the design
DGX / OVXJetson Orin / ThorH100 / HGXOmniverseDeepStream / MetropolisTensorRTTritonNVIDIA AI EnterpriseNeMo / NIM
Planned
Twin-to-work-plan optimisation · planned
Why it needs a GPU
Large as-built train twins, point clouds, CAD matching, simulation and multi-camera inspection are GPU-heavy by design. CPU-only cannot keep the twin current with the line.
MVP scope and first metric
Shadow mode on one marriage or fit-out area: prove fit and interface risk prediction against rework records, then graduate to supervised alignment guidance and twin-ranked build plans. First metric: marriage rework and fit-out rework on a named area.
Tier. Process tier, $16,000 per final-assembly area per month, or the anchor of the $120,000 per month Plant tier.
Open questions
- Per-plant sensor coverage, OT access and acceptance-data availability vary
- Cross-plant model transfer waits on customer IP and data agreements
Taktos · product family colour #178A47
Taktos
Sequence railcar production, JIS feeding, AMRs, commissioning and dynamic tests on takt.
The problem
Rolling-stock plants run long-cycle, variant-rich production where one missing harness, curing carbody, late bogie, unavailable brake-test slot or dynamic-test conflict can break delivery cadence. Schedulers juggle JIS feeding, buffers, AMR and AGV moves, station skills, rework, commissioning and customer acceptance with limited real-time prediction.
What Taktos does
Taktos is the line-and-logistics agent. It uses cuOpt, plant telemetry, work-package status, AMR and AGV state, inventory, coating cure windows, rework risk and commissioning resources to sequence work, feed kits, manage buffers and schedule brake tests, traction tests, ETCS, ERTMS or PTC commissioning, dynamic ride-quality runs and haul tests.
Who it is for
Rolling-stock plant directors, production-control teams, industrial engineers, logistics managers, commissioning managers and multi-site manufacturing networks.
Key features
- Line sequencing and takt-risk prediction for variant-rich rolling-stock builds
- JIS kit feeding, AMR and AGV routing, buffer and constraint management
- Coatyx flash and cure windows and Fitura train-twin readiness as constraints
- Commissioning and dynamic-test scheduling for brake, traction, signaling, ride-quality and haul tests
- Network analytics for first-time-through, PPM, delivery risk and capacity
NVIDIA stack
Named in the product brief as the design
H100 / HGXJetson Orin / ThorcuOptRAPIDSTritonNIMOmniverseNeMo
Planned
DGX or OVX for plant-twin scenario planning · plannedClosed-loop schedule auto-release with human gates · planned
Why it needs a GPU
Constrained optimisation over thousands of tasks, assets, variants, rework states and test slots is a GPU-native scheduling problem at network scale.
MVP scope and first metric
Shadow mode on one plant's sequencing: compare cuOpt plans against the released schedule and measure predicted takt risk against what happened, then graduate to recommended sequences the production-control team releases. First metric: line stoppage minutes and test-slot conflicts.
Tier. Process tier, $16,000 per month for line and logistics, often expanding to the $120,000 per month Plant tier and Enterprise network optimisation.
Open questions
- Per-plant sensor coverage, OT access and acceptance-data availability vary
- Closed-loop schedule release is planned, not built
Compare
What is in each tier
Products are licensed per process or line area on the Process tier and bundled on Plant and Enterprise.
| Capability | Process | Plant | Enterprise |
|---|---|---|---|
| Processes covered | 1 | All on site | All in network |
| Agent modules | 1 of 5 products | All agents | All agents + custom models |
| Train-and-plant twin (Fitura) | No | Included | Custom program twins |
| Plant-edge node | Included | Per process | Fleet managed |
| Guarded writes | Scoped tools | All scoped tools | Network policy |
| Car genealogy | Included | Included | Included + assurance packets |
| Autonomy levels | L1 to L2 | L1 to L3 | L1 to L4 when accepted |
| SSO and RBAC | No | Planned | Planned, with approval chains |
| On-prem deployment | No | Optional | Included |
| Support | Business hours | 24/7 | 24/7 + named engineer |
| Indicative price | $16,000/mo | $120,000/mo | $1M to $15M ACV |
Run timeline · scenario
The five products in one car
Every Railira run is an ordered, inspectable sequence. This is run_9b4e27: Car 07 of a 4-car metro consist · shell line B · bogie shop · booth 2 · marriage station. Scenario completes in 10 steps with one approval gate. Shell geometry held inside tolerance after a twin-ranked weld sequence, one heat-input change approved by the shift weld engineer, two torque traces routed to rework before brake test, the livery cure window protected in the sequence, marriage interface inside tolerance, car released with full genealogy.
run_9b4e27
car 07 · 4-car metro consist
illustrative scenario
SUCCEEDED
-
01
ingest.order
SUCCEEDED
0.8 s
Plant Orchestrator · Pulled the work packages for car 07 from MES and PLM: aluminium extrusion shell, EN 15085 weld class per joint, control-plan geometry tolerances, bogie torque specifications, livery recipe and cure window, fit-out and marriage sequence. Locked the approved control-plan envelope for the shift.
-
02
takt.sequence
SUCCEEDED
41 s
Line-and-Logistics · Sequenced the shift against live readiness: booth 2 cure window, bogie B-14 readiness, brake-test slot at 13:40, AMR fleet state and kit inventory. The plan moves car 07 marriage behind car 06 to keep the cure window and the test slot.
-
03
twin.simulate
SUCCEEDED
58 s
Train Twin · Simulated 18 candidate weld sequences for the roof-to-sidewall joints on the as-built twin and ranked them on predicted distortion, door-aperture geometry and heat input. Sequence #6 keeps the aperture inside tolerance with 2% less heat input than the standing sequence.
-
04
weld.sequence
SUCCEEDED
3 m 12 s
Weld-and-Structure · Recommended sequence #6 and the matching clamping plan to the shell line. Seam vision and laser-tracker geometry confirmed the roof-to-sidewall joints inside the control-plan tolerance on the first four seams.
-
05
approve.human
APPROVAL
52 s
Plant Orchestrator · Sequence #6 lowers heat input on the roof joints 2% below the standing weld procedure. Heat-input changes stay with a named engineer, so the step was held for the shift weld engineer and approved at 06:12:44 with the twin evidence attached.
-
06
bogie.torque
SUCCEEDED
continuous
Bogie-and-Running-Gear · Press-fit curves and torque traces on bogie B-14 read at the station. Two traction-motor mounting traces fell outside the expected profile and were routed to rework before brake test; the remaining safety-critical joints verified to spec.
-
07
coat.cure
SUCCEEDED
continuous
Coat-and-Livery · Booth 2 humidity and temperature, film thickness and flash time watched through the livery pass. One masking misalignment on the door band flagged by booth vision before the second coat; the cure window held and handed to the sequence as a hard constraint.
-
08
marry.align
SUCCEEDED
2 m 05 s
Marry-and-Couple · Carbody and bogie scans compared against target interface geometry before the lift. Fit-out readiness confirmed by Fitout-and-Interior for doors and glazing. Marriage guided inside tolerance; coupler and gangway checks verified and written to the twin.
-
09
inspect.fuse
SUCCEEDED
continuous
Quality-and-Anomaly · Weld and coating vision, in-line metrology, torque, wiring continuity and brake-test results fused at the edge across the car. Every finding localised to station, work package and shift, with the two torque reworks and the masking flag already closed.
-
10
car.release
SUCCEEDED
1 m 30 s
Conformance · Car 07 released: genealogy written (weld records per joint, geometry scans, torque traces, coating recipe and cure record, fit-out checks, marriage interface, approvals, test results) and dispositioned against the control plan and IRIS / ISO/TS 22163 criteria with cited evidence for the quality engineer's sign-off.
Accelerated computing
Physical AI, at the station
A seam is welded in seconds and a booth cannot wait for the cloud. Perception has to be local, deterministic and fast, so Railira runs GPU inference at the plant edge and keeps training, simulation and optimisation in the cloud or on-prem.
Edge perception
8 to 32 cameras per process area plus metrology, torque, thermal, robot and booth-instrument streams fused at the plant edge. Sub-50 ms weld and seam alerts and sub-150 ms coating and metrology prompts are the design targets. Jetson Thor is planned for higher sensor density.
Model serving
Weld-defect, geometry-risk, coating-risk, point-cloud and takt-risk models served at the plant edge with per-model versioning and rollback, routed by latency, tenant, sensitivity and cost.
Training
Weld and coating defect vision, metrology classification, control-plan reasoning and FEA-surrogate models fine-tuned on plant data and engineer corrections. 100 to 500 GPU-hours per plant during onboarding is the design target; dedicated capacity is planned.
Train-and-plant twin
As-built carbody, coating, marriage and line-flow twin. Weld distortion, geometry build-up, robot reach, coating coverage and layout simulated before the shift. GPU FEA and point-cloud metrology are the workloads; OVX scale is planned.
Synthetic data
Rare weld defects, coating contamination, geometry deviations, occluded seams and torque-escape cases generated and validated against real inspection distributions before promotion. Planned.
Optimisation
Line sequencing, JIS feeding, AMR and AGV routing, buffer management and commissioning-slot scheduling against live readiness.
Robot cells
Robotic weld, coating and assembly cells and AMR or AGV flow; path variants validated in simulation before release to the floor.
GPU ETL
Telemetry features from weld parameters, metrology, coating readings, torque, non-conformance and takt history for retraining, drift detection and network analytics.
Pricing
Priced per process, per plant, per network
Land on one workflow with a measured baseline. Expand by process, module and autonomy level. Outcome components are available on rework, line uptime, first-time-through and takt. Annual prepay carries a 15 to 20% discount.
Carbody, coating or assembly autonomy or inspection AI for one process. For plants validating ROI on the wedge.
- One process or line area, one agent module
- Plant-edge inference node included
- Controller and sensor reads, guarded recommendations
- Run log, audit trail, car genealogy
- Business-hours support
The whole plant: weld and structure, carbody and join, geometry and framing, coat and livery, marry and couple, fit-out and interior, quality and anomaly, line and logistics, and the twin.
- All agents on every process
- Fused inspection on every line
- Quality, conformance, line and takt
- Robotic cell and AMR supervision
- As-built train-and-plant twin
- 24/7 support, quarterly model review
Multi-plant OEM and tier-1 networks standardising on Railira, with custom carbody, coating, torque and takt models and a plant-edge fleet.
- Multi-plant fleet management
- Custom carbody, coating, torque and takt models
- Plant-edge fleet with on-prem option
- Group autonomy policy and approval chains
- SLAs, SSO and RBAC, dedicated engineer
FAQ
Straight answers
The questions plant directors, body and coating engineers and quality leads ask in the first meeting.
Only within an explicit tool allow-list with per-tool rate and magnitude limits, inside the approved control-plan envelope, and only at the autonomy level your plant has set. Design partners start in shadow mode, where Railira writes nothing. Heat-input and weld-sequence changes, schedule release and car release stay with a named approver.
Control returns to the robot, PLC or station's last known-good state. Railira is a supervisory layer on top of the controllers, safety systems and interlocks you already run, never a replacement for them, so an outage degrades the line to its current way of running, not to a stop.
Recommendations start from the control plan, the work instructions and the twin in the first weeks. Weld, geometry, coating and torque prediction improve as line history and labelled outcomes accumulate per car type. Onboarding is sized at 100 to 500 GPU-hours per plant and 50,000 to 250,000 synthetic images or scan variants per program, all design targets.
Only if you choose cloud training. Designs, control plans, program data, process models and quality records are isolated per tenant and never used to train another customer's models. Fully on-prem deployment at the plant edge is available for regulated plants, and cross-plant learning is planned and waits on data-rights agreements.
You are, the same as with any control strategy, which is why every write is policy-checked, bounded, logged and reversible, and why anything above your risk threshold waits for a named approver. The audit log records the request, the evidence, the envelope applied and the human decision.
A paid pilot on one process or line area, scoped to a single wedge with a success metric written into the order form: weld or coating rework reduction, geometry escapes, torque escapes, line stoppage minutes or first-time-through. It starts in shadow mode against a baseline from your own records, moves to assist mode when the engineer is satisfied, and expands only after the number moves.
Get started
Pick your first product
Start with one carbody, bogie, coating, fit-out or sequencing process and one measurable baseline. A paid pilot on weld and geometry, torque assurance, livery defects or line sequencing shows the number before you commit further.