AeroEstimate is not a “magic AI guesser”. It is a hybrid system that combines AI for damage interpretation with a structured cost model for parts, labor, paint and regional rates.
1. What AeroEstimate does
From a single photo or short video walk-around, AeroEstimate:
- Detects the damaged zones (front, rear, side, glass, roof, etc.).
- Identifies which parts are likely affected.
- Suggests the likely operation (repair vs replace vs R&I).
- Estimates labor, parts and paint using internal tables.
- Computes a total estimate and a realistic price range.
- Assesses drivability and urgency (can wait / fix soon / ASAP).
- Generates a clear summary for the vehicle owner or fleet.
2. AI’s role: interpretation, not pricing
AeroEstimate uses computer vision and large language models (OpenAI) to interpret what is damaged and how:
- Area of impact and damaged zones.
- Likely damaged parts and operations.
- Severity level (cosmetic, moderate, structural).
- Drivability and safety risk notes.
- Plain-language explanations.
AI is not used to invent prices, hours or totals. All numeric values come from AeroEstimate's internal cost model.
3. Cost model: tables, not prompts
The cost model relies on internal tables, not on GPT outputs. For each vehicle segment and panel type, AeroEstimate maintains:
- Typical OEM part price ranges.
- Typical labor hours for repair / replace / R&I.
- Paint and refinish hours per panel.
- Blend / adjacent refinish factors.
- Regional labor rates (e.g. Montreal, Toronto).
These tables are inspired by professional collision estimating practices (I-CAR, Mitchell, CCC, Audatex / Solera) and market data, then simplified and normalized for consistency.
4. Severity & adjustment factors
To avoid arbitrary numbers, AeroEstimate applies simple, documented adjustments based on severity:
- Cosmetic damage → base table values.
- Moderate damage → moderate uplift vs. cosmetic.
- Structural / high-risk zones (pillars, rails, roof, major front/rear) → higher uplift to reflect real-world complexity.
These factors are inspired by refinishing and complexity multipliers used in professional systems, with the goal of being predictable and explainable.
5. Regional labor rates & currency
Labor cost is computed as: labor hours × regional rate. Regional rates are based on typical body shop rates in that area and are periodically reviewed. AeroEstimate currently supports USD and CAD, with conversions handled via a single, consistent FX rate.
6. Price range instead of a single “magic number”
In the real world, two shops may legitimately choose different repair strategies (repair vs replace, OEM vs aftermarket, blend decisions). Hidden damage can also appear during teardown.
AeroEstimate reflects this by showing a range around the central estimate, representing:
- Different but reasonable shop decisions.
- Potential hidden damage.
- Differences in paint and blend operations.
7. Consistency & reproducibility
For the same vehicle, same images and same location, AeroEstimate aims to produce the same result, as long as the underlying model version and tables do not change.
The damage interpretation is normalized (e.g. “front bumper cover”, “front fascia” → front_bumper), and all pricing uses the same tables. GPT never manipulates the final numeric values.
8. Limitations & proper use
AeroEstimate is a decision support tool, not a legal or contractual estimate. It does not replace a full in-shop teardown and estimate by a certified appraiser.
It works best for:
- Vehicle owners who need a realistic idea of repair cost.
- Fleets that need to triage and prioritize repairs.
- Shops and partners who want faster pre-screening.
9. Questions & partnerships
For methodology questions, audits or partnerships, contact:
support@aeroestimate.app