[AUTOSAR] Automated Test & Validation Pipeline with API

AUTOSAR

RealThings’ proprietary AUTOSAR ARXML parsing tools

Automated test and Validation Pipeline with API

As automotive and IoT systems evolve, the number and variety of peripherals continue to grow, along with increasing dependencies and complex test coverage requirements. Consequently, developers—especially in large-scale projects—face significant challenges in executing fast and comprehensive testing cycles.

To address this, RealThings has implemented an automated test and validation pipeline that integrates everything from source code builds to various testing tools and formats. This end-to-end automation has enabled reductions in test time and costs by up to 95%.

A key strength of RealThings lies in its proprietary AUTOSAR ARXML parsing tools, including:

ECU & DIAG Extract Parsing Tool
- Supports ECU Extract schema versions 00044, 00046, 00049, and DEXT version 00047
- Parses detailed data based on communication protocols, messages, signals, values, and more
ECU Extract Delta Analysis Tool
- Performs delta analysis between two ECU Extracts
- Detects data modifications efficiently

These tools generate in-depth HTML reports with graphical diagrams to support detailed analysis. Together, they deliver powerful automation capabilities specifically tailored for AUTOSAR-based vehicle software development and validation.

Case Studies

  
        OEM 1 – ECU 1
        OEM 1 – ECU 2
        Re-Integration
      
        Explanation
        First ECU integrated and tested
        Second ECU integrated and tested
        Re-Integration of same software module developed by a different supplier
      
        Cost Savings
        
          Investment: € 230.000
          Found errors early saved approx € 700.000 in first test run.
        
          Investment: 0
          Found errors early saved approx € 700.000 in the second test run.
        
          Investment: € 5.000
          Find errors early saved approx $2.000.000 in the first test run.
        
        SpeedTest creation
        25% faster
        25% faster
        100% faster
      
        SpeedTest execution
        95% faster
        95% faster
        95% faster
      
        QualityDefect Escape Rate. (Percentage of defects that escape from one stage of development to the next)
        —
        —
        < 5%
      
        FlexibilityAre there restrictions on the software implementation method?
        Any software implementation can be validated to meet the quality and test criteria.
        Any software implementation can be validated to meet the quality and test criteria.
        Any software implementation can be validated to meet the quality and test criteria.
      
        Re-UsabilityCan the system be used for different ECUs, different implementations and across different vehicles?
        100%
        100%
        100%
      
        Result
        The positive business case and project impact for quality assured and automated configuration, integration and testing is very clear.

	OEM 1 – ECU 1	OEM 1 – ECU 2	Re-Integration
Explanation	First ECU integrated and tested	Second ECU integrated and tested	Re-Integration of same software module developed by a different supplier
Cost Savings	Investment: € 230.000 Found errors early saved approx € 700.000 in first test run.	Investment: 0 Found errors early saved approx € 700.000 in the second test run.	Investment: € 5.000 Find errors early saved approx $2.000.000 in the first test run.
SpeedTest creation	25% faster	25% faster	100% faster
SpeedTest execution	95% faster	95% faster	95% faster
QualityDefect Escape Rate. (Percentage of defects that escape from one stage of development to the next)	—	—	< 5%
FlexibilityAre there restrictions on the software implementation method?	Any software implementation can be validated to meet the quality and test criteria.	Any software implementation can be validated to meet the quality and test criteria.	Any software implementation can be validated to meet the quality and test criteria.
Re-UsabilityCan the system be used for different ECUs, different implementations and across different vehicles?	100%	100%	100%

Result	The positive business case and project impact for quality assured and automated configuration, integration and testing is very clear.

CI/CD deployed on AWS cloud

Infrastructure is built on Amazon EKS within a secure VPC
Divided into public and private subnets

Standalone AWS instance:

Hosts: Grafana
Monitors the cluster, applications and individual nodes
Monitors the Jenkins pipelines

Local server connected to hardware setup:

Internal server
Windows based
Hosts: RAFT and ECU testing tools
Connected to hardware for automated testing stages

AWS Node Group 1:

Public subnet
Linux based
Hosts: Jenkins, JFrog Artifactory, and Prometheus
Elastic IP accessible publicly
Integrated with Bitbucket via an Application Load Balancer
Autoscaling enabled for higher workload demands

AWS Node Group 2:

Private subnet
Windows based
Connected to the internet via a NAT Gateway for secure outbound traffic
Amazon S3 stores logs and backup data
Autoscaling enabled for build nodes to handle multiple pipelines parallelly