steddi: confidence in every single weld

• Reading values is subjective — it depends on light, viewing angle and eyesight.
• Slow and physically awkward; time pressure inside a track possession breeds shortcuts.
• Quality tracks the training, experience and diligence of the individual welder, so consistency between crews is poor.
• Misaligned rail ends are hard to grind out; localised peaks and dips persist in the track, producing dipped welds, impact loading and downstream tamping complaints.
• No built-in proof of work — compliance with work instructions and tender specifications is asserted rather than evidenced.

• Laser sensors read both rail ends simultaneously and instantly — vertical on all models, plus lateral on ONE.
• The values displayed are the same numbers a one-metre straight edge would give, so no target values need relearning.
• Works in full daylight; the LED display stays legible from a distance in bright sun.
• Magnetic feet hold the instrument on the rail head throughout the alignment.
• Proof of alignment through a geotagged photograph of the display — compliance becomes evidence.
• Explicit benefit claim: correct alignment with less dependency on training, experience or diligence.

• EntityBartoleni Engineering & Consulting, Terrigal NSW 2260, Australia · ABN 62 736 388 281. A small, focused engineering business operating at early launch stage, covering product design, the calibration chain, customer relationships and go-to-market from a narrow base.
• Proof pointsPublished customers span Queensland Rail (metro and passenger), Rio Tinto (heavy-haul mine rail) and Sydney Trains / TfNSW (urban). That mix — urban passenger, regional and private heavy-haul — covers the three distinct procurement worlds in Australian rail. European norms (EN 14730-1) are already referenced in the company's material.
• Calibration annuityEvery unit ships with a certificate calibrated against a grade-A master straight edge per AS 1003-1971, with annual recalibration advised per AS 1085.20. Customers can self-calibrate using offered jigs — a pragmatic trust-building move that keeps device cost down while making the relationship recurring.

• Hardware sale plus a calibration annuity. Rugged IP65 tools with a defined annual calibration event is a classic instrument-business model — compare survey instruments or torque calibration. Margins hold as long as the metrology chain stays credible.
• The app is a companion, not a platform. GO's app performs logging, analysis and export per device. There is no visible cloud, team or network tier — no fleet dashboard, no track-kilometre view, no trend analytics. That is the gap the roadmap addresses.
• Compliance is the wedge into conservative buyers. Full compliance with work instructions and tender specifications, backed by geotagged photographic proof, speaks directly to how rail principals actually audit contractors today.
• The training-reduction claim is a labour-arbitrage play. Rail welding crews are scarce and ageing. A device that delivers correct alignment irrespective of the operator's level of training addresses a real workforce constraint — and pairs naturally with simulation-based training.

• Real deployments with three tier-one rail organisations — an uncommon achievement for a business of this size.
• A genuine first-mover claim in pre-weld alignment, rather than another post-weld inspection tool.
• Metrology credibility: traceable calibration, standards-native to both AS and EN norms.
• Products designed with obvious field empathy — magnetic feet, sunlight-readable LEDs, standard Bosch batteries, 160 °C short-term tolerance.
• Small, coherent line: three SKUs, one workflow, no bloat.

• Concentration risk across hardware, firmware, app, calibration and sales in a very small organisation.
• The data story stops at the phone — no cloud or platform tier is visible.
• Go-to-market polish lags the product; the web presence is thinner than the engineering.
• No published integration story with welding machines, grinding trains or track-recording systems.
• Brand surface is split across steddi and Bartoleni, across two domains.

• A weld data platform — becoming the system of record for weld quality per network is the durable position.
• European expansion — EN 14730-1 is already referenced, and EU rail welding volume is large and standards-driven.
• Automated welding QA — every new flash-butt machine and grinding robot needs independent geometry verification; see the auto welding review.
• Remote-corridor workflows — Pilbara and regional Queensland deployments need offline-first capture, and a heavy-haul customer is already on board.
• Training simulators — the device defines what "correct" is; a simulator can teach to it.

• Incumbent measurement vendors adding a pre-weld mode to established digital straight edges.
• Track-recording cars and inspection trains absorbing weld geometry into their own datasets.
• Low-cost copycat devices once the concept is proven — the supplier base for rail welding hardware is deep.
• Rail procurement cycles are slow, and a small vendor can simply be outwaited.
• Calibration credibility is a single point of attack: if a network disputes the metrology chain, the premise wobbles.

• Gap 1 · No data platformMeasurements live on phones. Networks buy trends and assurance, not single readings. Without aggregation, a precision instrument stays a tool rather than infrastructure.
  FIT → chanalyse pipeline — time-series store, anomaly detection, automated publishing
• Gap 2 · No spatial viewNo map or 3D context for welds. Track engineers reason in kilometres and corridors. Geotagged photos without a map are half-evidence.
  FIT → Rule the City terrain engine — real DEM and OSM data, already renders corridors
• Gap 3 · Remote-area workflowHeavy-haul corridors have no reception. Bluetooth-to-phone assumes someone later syncs. Offline-first capture with store-and-forward comms is unsolved.
  FIT → RFAI — offline-first field intelligence stack with no cloud dependency
• Gap 4 · No automation interfaceThe device is human-only. Automated welding plant and grinding robots are arriving; they need machine-readable QA and, eventually, supervised teleoperation.
  FIT → robot_hand teleoperation · on-device vision models

Analysis of public information only — steddi.com.au product pages, published FAQ material and referenced standards. August 2026.