Per NTT Data’s Insurtech 2025 report, telematics-augmented underwriting is projected to lift pricing accuracy by roughly 20% and cut loss ratios by roughly 15% — a sector-wide estimate for UBI generally, not an Arctura-measured result. UBI reads how a car is driven. Arctura adds an A–F rating for what the asset is — an overlay on top of your UBI and tariff model, not a replacement.
global EV-insurance pool by 2030 — external industry projection (GlobeNewswire press release), cited illustratively.
EV vs ICE repair in Europe (Insurance Europe) — a sector figure, not an Arctura result.
from dynamic telematics / UBI, per NTT Data 2025 — a sector estimate, not an Arctura-measured result.
UBI programs report ~15% loss-ratio improvement + up to 30% retention (NTT Data) — third-party benchmark, not Arctura's.
For insurers, Arctura is a recognition layer — not a data vendor you buy a feed from. In our commercial model the asset’s operator or fund is the paying party for the A–F rating; an insurer can then recognise that rating in the terms it offers, treating a strong grade as a risk-reducing measure — much as it would reflect any credible risk-reduction evidence in the terms it writes.
The party that owns the risk-reducing measure commissions and pays for the rating on the fleet it owns or operates. Volume-based pricing scales with portfolio size, agreed per engagement.
A strong A–F Arctura rating can support better terms — they are not billed for the data. Better terms are a possible outcome of clearer evidence, never a guarantee.
A pre-close grade and degradation envelope on the asset behind the loan, grounded in a validated model — not a spreadsheet assumption.
Pricing is agreed with the operator or fund — see pricing →
UBI tells you how the car is driven. TAS tells you what the asset is. Industry research (NTT Data) projects ~20% pricing accuracy, ~15% loss-ratio and ~30% retention gains from UBI generally — a sector estimate, not an incremental Arctura adds. We sit on top of UBI, not against it.
TAS is a tariff-pricing input, not a Solvency II internal-model input. Whether a given standard applies is for the adopting insurer's own actuarial and compliance functions to determine. AI Act: a self-assessment is planned (targeting Q4/2026).
You process the score, not the raw telemetry. EU-sovereign by construction, Hetzner Helsinki residency, ICT-provider terms designed to align with DORA, sub-processor list in the DD pack.
The scoring methodology and its input pillars are public; the pillar weights and internal accuracy metrics are trade-secret — the same standard a rating agency uses when it does not publish its sovereign-rating weighting formula.
Same as a rating agency not publishing its sovereign-rating weighting formula.
How a pilot works: an operator or lessor commissions a scoped sample of its assets over six to eight weeks, and Arctura returns an independent A–F rating with a recomputable, hash-verified evidence trail for that rating. Insurers recognise that rating in the terms they offer; they are not billed for a feed.
ROADMAP · NOT YET CONTRACTUALLY OFFERED
ICT-provider terms and incident-reporting designed to align with DORA. EU-resident.
An operator or lessor commissions a scoped sample of its assets; Arctura returns an independent A–F rating with a recomputable, hash-verified evidence trail. Insurers recognise the rating — they are not billed for a feed.