Author: i2cat

CARAMEL 2nd OEM & Partner Workshop

CARAMEL 2nd OEM & Partner Workshop

The CARAMEL project will held the second OEM and Partners workshop on 16 November 2021, from 14:00 to 17:00. This workshop aims to present the most outstanding results of the research and development carried out by the different members of the consortium. Furthermore, this workshop is also oriented to OEM representatives, which could increase the visibility of the project and collect opinions on the topics addressed by CARAMEL.

The objective of this workshop will be to highlight the achieved results toward the development of Artificial Intelligence-based cybersecurity for connected and automated vehicles.

The CARAMEL project bases its research on 4 fundamental pillars for the safety of connected autonomous vehicles.

Main Pillars of the project:

Pillar 1: Autonomous Mobility
Pillar 2: Connected mobility
Pillar 3: Electromobility
Pillar 4: Remote Control Vehicle (RCV)

Autonomous Mobility

  • Taxonomy of the attacks.
  • Cyber attacks Detection and Mitigation on Sensing and Navigation modalities.
  • Elevation of Perception Engines as core modules for cyber-attack detection  & mitigation engines.
    • Towards robustifying D-CNNs to tackle adversarial attacks on the scene layer.
  • Multimodality and Redundancy of Sensors beat malicious attacks on CAVs.
  • Fall back actions to enhance safety.
  • Key Performance Indicators for assessing Mitigation Performance.

Connected mobility

  • Interoperability between radio technologies for V2X communications
  • Secure V2X communications and related hardware
  • GNSS, V2X and HW attack detection and response process
  • Vehicle tracking using its signature certificates

Electromobility

  • Communication architecture of an EV charging network

Remote Control Vehicle (RCV)

  • ML-based traffic between 5G-RCV 
  • Control Center anomaly detection
  • prediction algorithm development

Don’t miss this excellent opportunity to learn more about the activities carried out to develop cybersecurity based on artificial intelligence for autonomous and connected vehicles.

CARAMEL at MMSP 2021

CARAMEL at MMSP 2021

The IEEE 23rd International Workshop on Multimedia Signal Processing, organised by the IEEE Signal Processing Society, Centre for Immersive Visual Technologies and Tampere University, will be held from 6th-8th October in Tampere, Finland.

The congress will be the meeting point for professionals from academia and industry developing or are carrying out projects in the field of multimedia signal processing, aiming to share knowledge, exchange ideas and explore future research directions.

The IEEE MMSP 2021 is a hybrid event. This includes the face-to-face gathering at Scandic Hotel Rosendahl and online engagement through a virtual platform.

This event will include the presence of the University of Patras members of the CARAMEL consortium, who will be presenting the paper:

Deep multi-modal data analysis and fusion for robust scene understanding in CAVs

Deep learning (DL) tends to be the integral part of Autonomous Vehicles (AVs). Therefore the development of scene analysis modules that are robust to various vulnerabilities such as adversarial inputs or cyber-attacks is becoming an imperative need for the future AV perception systems. In this paper, we deal with this issue by exploring the recent progress in Artificial Intelligence (AI) and Machine Learning (ML) to provide holistic situational awareness and eliminate the effect of the previous attacks on the scene analysis modules. We propose novel multi-modal approaches against which achieve robustness to adversarial attacks, by appropriately modifying the analysis Neural networks and by utilizing late fusion methods. More specifically, we propose a holistic approach by adding new layers to a 2D segmentation DL odel enhancing its robustness to adversarial noise. Then, a novel late fusion technique has been applied, by extracting direct features from the 3D space and project them into the 2D segmented space for identifying inconsistencies. Extensive evaluation studies using the KITTI odometry dataset provide promising performance results under various types of noise.

7th October 2021

The IEEE 23rd International Workshop on Multimedia Signal Processing

CARAMEL at AIBIGDATA21

CARAMEL at AIBIGDATA21

The AI & Big Data Congress 2021, organised by the Innovation Centre for Data Tech and Artificial Intelligence, will be held on 15th September at the AXA Auditorium in Barcelona.

The congress will be the meeting point for professionals, suppliers, and companies developing or are carrying out projects in the field of AI & Big Data.  

This edition will talk about the new Artificial Intelligence challenges, market trends and best practices of pioneer companies, technological innovations and their applications; and success stories explained in detail. 

Don’t miss the opportunity to participate in this event in which Jordi Guijarro (CARAMEL’s Project coordinator) will present a session of the proof of concept envisioned by the CARAMEL project.

H2020 CARAMEL: AI-based cybersecurity for connected and automated vehicles

15th September 2021

Registration link

7th edition AI & Big Data Congress | 14-15 September 2021 (aicongress.barcelona)

Automotive Threat Modelling Tutorial

Automotive Threat Modelling Tutorial

CARAMEL aims to produce a cybersecurity system based on artificial intelligence to combat cyber threats present in autonomous vehicles.

However, have you ever wondered how to perform a detailed analysis to identify threats in autonomous vehicles?

CARAMEL’s members created a tutorial that explains how to develop and analyse an automotive threat model using Microsoft’s Threat Modelling tool through the STRIDE technique.

Threats
Spoofing
Tampering
Repudiation
Information disclosure
Denial of service
Elevation of privilege

The STRIDE technique tries to identify as many possible threats in the system by decomposing a more extensive system into its most relevant components.


The STRIDE technique was named after the threats, it can identify.

We invite you to register and download the tutorial at the following link.

References

Uncover Security Design Flaws Using The STRIDE Approach | Microsoft Docs

CARAMEL demo video series

CARAMEL demo video series

As part of the project’s dissemination activities, we have opted to share video demonstrations of some of CARAMEL’s developments.

Each of the videos is composed of an introduction to the developed system, followed by a general explanation of the system, and finally a presentation of the most outstanding results of the particular system.

Within the scope of the developments presented, you can find systems developed for some of the pillars addressed by Caramel such as:

Pillar 1: Autonomous vehicles
Pillar 2: Connected vehicles

The titles of the videos are

Title
RSU-OBU-TestBed
Traffic sign anomaly detection and mitigation pipeline
Detecting possible attacks on the camera sensor using a deep learning approach
In-vehicle Location Spoofing Attack Detection
Holistic Situational Awareness with ML Application
Collaborating mitigation mechanism against GPS spoofing

The videos can be found at the following link or through the following video gallery

CARAMEL Workshop

CARAMEL Workshop

As part of the CARAMEL activities on 27 May 2021, starting at 09:00, a 2-hour virtual workshop will be held. The CARAMEL project will introduce their activities to representatives of OEMs to increase the reach and gather opinions about the topics addressed by CARAMEL.

The objective of this workshop will be to highlight the achieved results toward the development of Artificial Intelligence-based cybersecurity for connected and automated vehicles.

Main Pillars of the project

Pillar 1: Autonomous Mobility
Pillar 2: Connected mobility
Pillar 3: Electromobility
Pillar 4: Remote Control Vehicle (RCV)

Registration Link

https://us02web.zoom.us/meeting/register/tZcrd-qoqjMvHNLMw6lN3bhzAkkMwaMXky1U

Korean Partners contribution plans to CARAMEL

Korean Partners contribution plans to CARAMEL

With the inclusion of the Korean (KR) Partners i.e. KATECH Korea Automotive Technology Institute, MOBIGEN and ETRI The Electronics and Telecommunications Research Institute to the CARAMEL consortium, there is now a 4th pillar ‘Remote Control Vehicle (RCV)’ added to CARAMEL pillars i.e. Pillar 1: Autonomous Mobility, Pillar 2: Connected Mobility & Pillar 3: Electromobility. The focus of the KR partners would be to develop a vehicle gateway to remote control vehicle and built an Automotive Cyber Security module based on AI edge. The plan is to have ML based intrusion detection and estimation algorithm in the Gateway & RCV controller.

The activities planned by the partners around the 4th Pillar are:

  1. KATECH:
    • With support from Renault Samsung Motors will provide Remote Controlled Vehicle (Renault Arkana) with remote control functionality (front, left, right and rear view streaming system, Around View Monitoring System).
    • Real time dashboard to monitor vehicle status and ADAS system status(Land Departure Warning, Blind Spot Warning, Forward Collision Warning, Parking Assist).
    • Urban type proving ground for conducting PoC.
    • Scenario driven attack and pilot
    • Provide RCV HW platform
    • Lead dissemination activities for KR partners and contribute to CARAMEL dissemination.
  2. MOBIGEN:
    • Analysis of security & privacy requirements for RCV.
    • Threat modelling for RCV.
    • Investigate CyberThreat detection and response techniques for RCV.
    • Provide big data analysis HW & SW platform for CyberThreat detection (AI Edge) at the vehicle (front end)
    • Scenario driven attack and pilot
    • Work with Korean vehicle manufacturers and cybersecurity SW vendors for Business model in Korea and collaborate with EU partners on any potential global BM development.
  3. ETRI:
    • Analysis of security & privacy requirements for RCV.
    • Threat modelling for RCV.
    • Investigate CyberThreat detection and response techniques for RCV.
    • Provide ML based CyberThreat detection and analysis SW at the control center of the RCV (Back end).
    • Scenario driven attack and pilot
    • Collaborate with EU standardization partners to co-edit contributions and reflect onto relevant standards.

The KR partners would be utilizing the EU related deliverables as base references and then identify and investigate additional / unique security requirement as well as threat models with regards to RCV.

We look forward to the contributions and activities along with the Korean Partners.

CARAMEL EC review meeting

CARAMEL EC review meeting

On 24 February 2021, the CARAMEL project held its first review meeting aiming to evaluate the performance during the first half of the project’s life.
The meeting was attended by representatives of the European Commission who acted as evaluators, the CARAMEL project officer and representatives of each of the CARAMEL consortium member companies.
Presentations relevant to each of the work packages were made during the meeting. In which the work carried out in the following areas can be highlighted.

  • Mitigating Image Noise Attacks with Deep Learning
  • Noise Suppression Using Total Variation PDE
  • Cyberthreat Detection and Response Techniques for Cooperative Automated Vehicles
  • Cyberthreat Detection and Response Techniques for Plug-in Electrical Vehicles
  • DriveGuard: Robustification of Automated Driving Systems with Spatio-Temporal Convolutional Autoencoder
  • Mitigation on camera sensor attacks using LiDAR
  • In-vehicle Location Spoofing Attack Detection
  • GPS Spoofing Attack in Collaborative Vehicles
  • Backend Solution
Joint standardisation workshop on cyberattack projects

Joint standardisation workshop on cyberattack projects

On 22 January 2021, CARAMEL will be presented by our technical coordinator Peter Hofmann as part of the first joint standardisation workshop on cyberattack projects. It will be attended by other consortia that are also funded by the H2020 initiative.

All participating consortia belongs to the H2020-SU-ICT-2018-2020 call funded by the H2020 initiative which funds cybersecurity related projects.

CARAMEL is pleased to participate in this initiative. We will be updating you shortly on the joint activities arising from the workshop.

5th Innovation & Entrepreneurship Forum (IEF2020) 2nd Presentation

5th Innovation & Entrepreneurship Forum (IEF2020) 2nd Presentation

IEF 2020 is organized by the Centre for Entrepreneurship of the University of Cyprus in collaboration with PwC Cyprus. The 2020 Forum will bring together AI experts, researchers and professionals, decision-makers, entrepreneurs and game-changers to discuss the Challenge of Artificial Intelligence and address the most defying questions around the future of AI and its impact on society, economy and politics.

Recently our collaborators from University of Cyprus presented their AI research results “DriveGuard: A deep learning technique for countering cyber-attacks against autonomous vehicle camera sub-systems in the context of the H2020 CARAMEL project” as part of the activities of the 5th innovation & Entrepreneurship Forum (IEF2020): “The Challenge of AI”

Abstract

In the context of CARAMEL, we investigate the effect of camera sensor attacks for visual AI tasks in autonomous vehicles and present research results for a deep learning technique called DriveGuard, that act as a defense and mitigation mechanism for autonomous vehicle perception systems. This approach can provide protection against scenarios where an attacker can gain access to vital vehicle components by taking advantage of over-the-air updates in order to instantiate attacks that are not immediately detectable or perceivable, in contrast to just switching components off. DriveGuard utilizes the convolutional autoencoder family of deep learning networks to build an approach for efficient image reconstruction and anomaly detection. The approach will be integrated into an embedded anti-hacking device that will be capable for passive detection of attacks on an autonomous vehicle’s visual perception modules.

Approach to guard the segmentation module. The algorithm identifies different objects on the scene.
Left) Original image, Center) Attacked image, Right) Reconstructed image

You can find more information about this topic at the following video:

Theme: Elation by Kaira.