⚡ electrical engineering

Highway Readiness Assessment for SAE Levels of Automation and V2X Notification

This paper proposes a quantitative Highway Readiness Index (HRI) to evaluate infrastructure support for SAE automation levels through expert-weighted static ODD attributes, demonstrates its application in a case study to identify infrastructure gaps, and outlines its integration into IVIM messages for communicating segment-level automation guidance to connected vehicles.

Original authors: Lorenzo Italiano, Federico Marino, Mattia Brambilla, Giovanni Megna, Benedetto Carambia, Monica Nicoli

Published 2026-05-26

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Lorenzo Italiano, Federico Marino, Mattia Brambilla, Giovanni Megna, Benedetto Carambia, Monica Nicoli

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to drive a self-driving car. You might think the car is the only thing that matters, but this paper argues that the road itself is just as important. The authors, a team from a university and a highway management company in Italy, realized that while we are building better cars, we don't have a good way to measure if our roads are actually "ready" for them.

Here is the breakdown of their solution, explained simply:

1. The Problem: The "Driver's License" Gap

Think of self-driving cars like students learning to drive.

SAE Levels 1–2 are like a student with a learner's permit: They can steer or brake automatically, but a human must always watch the road.
SAE Levels 3–4 are like a student with a full license: The car drives itself, and the human only takes over if something goes wrong.

The problem is that roads are like classrooms. Some classrooms have whiteboards, good lighting, and quiet desks (great for learning). Others have peeling paint, broken lights, and noisy construction next door (terrible for learning). Currently, road managers don't have a test to see if their "classroom" is good enough for a "student" to graduate to a higher level of driving.

2. The Solution: The "Highway Readiness Index" (HRI)

The authors created a scorecard called the Highway Readiness Index (HRI). Think of this as a report card for the road.

Instead of just looking at the road and guessing, they broke the road down into specific, measurable parts, like:

The Paint: Are the lane lines bright, wide, and consistent? (If the paint is faded, the car gets confused).
The Signs: Can the car read the signs clearly?
The Surface: Is the asphalt smooth, or is it bumpy and cracked?
The Digital Map: Does the car have a super-detailed "GPS map" (HD Map) loaded before it even starts driving?

3. How They Made the Scorecard

To make sure the scorecard was fair, they asked 17 experts (like professors, engineers, and tech directors) to vote on what matters most.

For the "Learner" (SAE 1–2): The experts said clear paint and visible signs are the most important. If the lines are messy, the human driver needs to take over.
For the "Graduate" (SAE 3–4): The experts said everything matters more, but HD Maps and emergency lanes became critical. A self-driving car needs a perfect digital map to know exactly where it is, and it needs a safe place to pull over if it gets stuck.

They turned these opinions into math weights. If a road has great paint but no HD map, the score for the "Graduate" level drops significantly.

4. The "Traffic Light" Message (IVIM)

Once they calculate the score, they turn it into a digital message sent from the road to the car. Imagine a traffic light, but instead of Red/Yellow/Green, it tells the car:

Green (High Score): "You are safe to drive yourself at Level 4."
Yellow (Medium Score): "You can try Level 2, but be careful."
Red (Low Score): "Do not trust the automation; the human must drive."

This message is sent via a standard system called IVIM. It's like the road whispering to the car, "Hey, the paint is peeling ahead, don't try to drive yourself here."

5. Testing it on a Real Highway

They tested this on a 24-kilometer stretch of highway in Italy.

Scenario A (Normal Day): The road was well-maintained. The score was high, and the system said, "Yes, cars can drive themselves here."
Scenario B (Road Work): They simulated road construction with messy temporary paint. The score dropped sharply. The system immediately told the car, "No, take over the wheel," because the conditions were too risky for a self-driving car.
Scenario C (Bad Maintenance): They simulated a road with cracked pavement and missing signs. The score crashed. The system said, "No automation allowed."

The Big Takeaway:
The paper shows that you can't just build a smart car and expect it to work everywhere. You need a smart road to match it. By using this "Report Card" (HRI), road managers can see exactly where they need to fix the paint or add digital maps to make the highway safe for self-driving cars. It turns the vague idea of "is this road ready?" into a clear, numerical answer.

Technical Summary: Highway Readiness Assessment for SAE Levels of Automation and V2X Notification

Problem Statement

Despite rapid advancements in Automated Driving (AuD) technologies, road operators currently lack practical, quantitative methods to assess whether their existing infrastructure can support specific Levels of Automation (LOA). While standards define Operational Design Domains (ODDs) and automation levels, they do not provide a reproducible framework to map static roadway characteristics to the operational requirements of Automated Driving Systems (ADSs). Consequently, authorities cannot systematically identify physical or maintenance-related factors that hinder higher automation levels, nor can they communicate infrastructure capabilities to Connected and Automated Vehicles (CAVs) in a standardized manner. Existing frameworks often focus on digital connectivity or rely on extensive fleet testing, failing to offer a scalable, infrastructure-centric metric for network-wide assessment.

Methodology

The paper proposes a quantitative Highway Readiness Index (HRI) designed to translate static and semi-static infrastructure conditions into measurable attributes weighted by their impact on ADS performance. The methodology is structured into three stages:

Attribute Operationalization and Weighting:
- Taxonomy: The authors define a set of static ODD attributes grouped into four macro-categories: Road Markings & Signage, Road Maintenance & Management, Roadway Design & Safety Features, and Preloaded High-Definition (HD) Maps.
- Expert Survey: To derive weights, a survey was conducted with 17 professionals from academia, industry, and infrastructure management. Respondents assessed the impact of attribute absence or degradation on two LOA groups: SAE 1–2 (Assisted Driving) and SAE 3–4 (AuD).
- Differentiation: The survey revealed that while lane visibility and maintenance are critical for SAE 1–2, attributes like HD maps, emergency lanes, and dedicated AV lanes become significantly more important for SAE 3–4.
Segment-Level Readiness Model:
- Road segments (e.g., 100 m) are evaluated against measurable criteria derived from national design standards.
- An observed value $v_i \in \{0, 1, 2\}$ is assigned to each attribute based on adequacy.
- A readiness score $S^{(\ell)}$ is calculated for a specific automation level $\ell$ using a weighted sum normalized by the maximum possible score:
  $S^{(\ell)} = \frac{\sum w_i^{(\ell)} \cdot v_i}{\sum w_i^{(\ell)} \cdot v_{max}} \times 100$
- Distinct weights ( $w_i^{(\ell)}$ ) are applied for SAE 1–2 and SAE 3–4 to reflect their differing sensitivities to infrastructure quality.
Infrastructure Guidance and IVIM Integration:
- Scores are classified into three categories: Unlikely to be suitable (0–33%), May be suitable (33–66%), and Highly likely to be suitable (66–100%).
- A conservative threshold of 66% is established. Only SAE levels with scores above this threshold are communicated to vehicles.
- These recommendations are embedded into the standardized Infrastructure-to-Vehicle Information Message (IVIM) within the Cooperative Intelligent Transport System (C-ITS) framework, allowing Road Side Units (RSUs) to broadcast segment-level automation guidance.

Key Results

The methodology was validated through a case study on the Gallarate–Gattico corridor (24 km) in Northern Italy, managed by Autostrade Per l'Italian (ASPI).

Baseline Conditions: Under current well-maintained conditions, the corridor consistently supports SAE 1–2 and SAE 3–4 (scores >66%), indicating that existing high-quality highways can support advanced automation in stable configurations.
Roadworks Scenario: Simulating temporary roadworks (km 11–17) with degraded markings and geometric transitions caused a sharp drop in readiness, particularly for SAE 3–4. Several segments fell below the 66% threshold, correctly triggering a conservative IVIM message that withheld automation recommendations for those sections.
Maintenance Scenario: Simulating long-term poor maintenance (degraded markings, pavement deterioration, missing safety devices) resulted in significant readiness drops across the entire corridor, with many segments falling into the "Unlikely to be suitable" range for both automation groups.
Sensitivity Analysis: A theoretical analysis demonstrated that Preloaded HD Maps have the highest weight for SAE 3–4. The inclusion of HD maps in degraded infrastructure scenarios partially compensated for physical deficiencies, shifting readiness scores from "Unlikely" to "May be suitable," highlighting the critical role of digital layers alongside physical infrastructure.

Significance and Claims

The paper claims to provide the first quantitative, infrastructure-centric, and scalable methodology to map static roadway characteristics to ADS operational requirements. Its primary contributions include:

A Reproducible Metric: The HRI offers road operators a structured approach to evaluate infrastructure readiness and identify specific gaps (e.g., marking consistency, vegetation control) that limit higher automation levels.
Differentiated Assessment: By distinguishing between SAE 1–2 and SAE 3–4, the framework acknowledges that higher automation levels rely more heavily on stable geometry, fallback areas, and digital redundancy (HD maps) rather than just visual cues.
Operational Integration: The work bridges the gap between infrastructure assessment and vehicle operation by defining how readiness scores can be directly translated into standardized V2X messages (IVIM). This enables infrastructure to proactively guide vehicles, potentially preventing disengagements by warning drivers or systems of unsuitable segments before they are encountered.
Strategic Planning: The framework serves not only as a diagnostic tool for specific corridors but also as a strategic instrument for network-wide planning, allowing authorities to prioritize maintenance and digitalization investments based on their impact on automation readiness.

The authors conclude that while the current framework focuses on static and semi-static conditions, it lays the groundwork for future integration with dynamic factors (traffic, weather) and expanded international validation.