OpenResearch
Yellow ADAC vehicle on a wet mountain road at dusk, with light trails and village lights in the valley

Case StudyADAC · Connected vehicle

The car delivers data. We make it usable.

ADAC Smart Connect, built and run by us since 2021. From the telematics dongle to manufacturer data that now reaches roadside assistance.

  • 99.97%Availability over 3 years of trial operation
  • 15MData transmissions a day
  • Since 2021Development and operations

8 min read

A connected car platform for ADAC. Built and operated since 2021.

OpenResearch built ADAC Smart Connect on Microsoft Azure from scratch and has been developing the platform ever since: the Smart Connect app for members, the SCOPE operations console and a backend that takes dongle and manufacturer data all the way to roadside assistance.

At a glance:

  1. Client

    ADAC

  2. Period

    Since July 2021

  3. Technology

    Azure, .NET 8, Terraform

01Starting point

No platform. Just a goal.

In July 2021, there was no platform, just the goal of making vehicle data useful for members and roadside assistance.

Modern vehicles and connected devices send data continuously. Turning that into services takes a platform that reliably receives and processes these events and passes them on to apps. It started with telematics dongles in the OBD port and an app for members. Today the data also comes directly from the vehicle manufacturer.

02Approach

Invent little. Know a lot.

Domain knowledge is often underestimated. It significantly reduces schedule risk and, in our view, is the cornerstone of successful projects.

We relied on dependable, forward-looking technologies we knew well, and on experience from connected-vehicle projects since 2016, from a complete scooter-sharing platform to event-driven processing of IoT device data.

Domain knowledge since 2016

Experience is the best schedule insurance.

Step by step since 2021:

  1. 2021

    Smart Connect kick-off

    Sprint 0 in July. Telematics dongles connected, Azure infrastructure, an operations panel for subscriptions and shipping. Production-ready by year end after a pentest.

  2. 2022

    Expansion into trial operation

    The operations panel becomes SCOPE. Trips with eco and safety scores, alarm mode, remote diagnostics from the app, GDPR services. Rollout to test users.

  3. 2023

    Automated diagnostics

    Members can start the Automated Member Diagnosis (AMD) in the app. Once all control units have been read, the result is available immediately. Data science work begins in the project. The proof of concept goes into trial operation.

  4. 2024

    Stable operation

    A fixed release, support and test process, .NET 8. Maintenance reminders, battery warnings, an explainable safety score. First manufacturer data is evaluated.

  5. 2025

    Manufacturer data instead of dongles

    OEM proof of concept 1.0 and 2.0 across several manufacturers. Onboarding of manufacturer connections in the app, backend and SCOPE. Data quality measures, another pentest.

  6. 2026

    Roadside assistance pilot

    Reactive use case pilot: the backend receives breakdown data, retrieves vehicle data and assigns it to the case. Handling takes place in SCOPE.

03Decisions

What we chose.

Four decisions: risk out, speed in.

Each one served the same goal, and each had an obvious alternative that we deliberately rejected.

What we chose, and why:

  1. Cloud

    Microsoft Azure

    The team knew Azure very well. Familiarity beats theory, and it still holds today.

  2. Infrastructure

    Terraform

    Instead of manual environments: repeatable and traceable.

  3. Data

    Lambda architecture

    Instead of separate systems: one data foundation, two speeds.

  4. Data source

    Open to manufacturers

    Instead of being tied to the dongle: vehicles also connect directly via the manufacturer interface.

Vehicles& devicesOne data foundationApps · Real timeBI · Batch
One data foundation. Two speeds.

04Solution

Real time for the apps. The same data for BI.

Vehicle data is processed the moment it arrives: on average more than 1,500 data points per vehicle per day.

Members see it in the Smart Connect app, and ADAC works with it in SCOPE. In parallel, it lands in long-term analytical storage, where data science looks for patterns for roadside assistance.

Three interfaces, one core:

  1. Smart Connect app

    For members

    Mileage, fuel, battery, trips and remote diagnostics with recommended actions.

  2. SCOPE

    For operations

    Subscriptions, shipping, devices, compatibility and case handling for roadside assistance.

  3. Backend on Azure

    For everything in between

    Dongle and manufacturer data, GDPR services, deletion and archiving concept.

ADAC Smart Connect app: diagnostic report with control units read, fault codes, battery and alternator status
SCOPE: ADAC roadside assistance case handling, with a list of breakdown cases and vehicle data; on the right, battery voltage, fault codes and likely cause
Result

Result

  1. 99.97%.Availability over three years of trial operation. The requirement was 98.5%.
  2. 15M.Data transmissions a day across 10,000 vehicles.
  3. 8 TB+.Vehicle data collected since 2021.
  4. 100M+.Files in long-term analytical storage.
01 / 04

06Collaboration

Who did what.

Clear roles from the start.

ADAC owns the product and brings domain expertise and roadside assistance; OpenResearch brings architecture, delivery and operations. One team since 2021.

OpenResearch

Architecture & delivery

ADAC

Product & domain

  • Product ownership
  • Target vision and functional requirements
  • Roadside assistance process and case handling

One team. Two roles.

A similar project?

From zero to platform, within the original timeframe. Where are you hitting the limit?

Philipp, CEO at OpenResearch
PhilippCEO

Your contact · Initial call · Free · online

30 minutes.
An honest assessment.

Philipp listens and tells you honestly what’s realistic.

What happens next:

01 · 30 min · freeInitial call

An honest assessment with Philipp. Book directly above.

02 · 2 hrs · freeIntro session

We work on your real question with the team that would later build it. You leave with a sketch of the possible way forward.

Request an intro session
03 · Fixed price · optionalWorkshop
04Project

If it becomes a project, the workshop fee is fully credited.

Delivered in this project with.

4 services, one team:

More case studies.

Real projects, real numbers:

FAQ

Good to know.

What is a connected-vehicle platform?

A connected-vehicle platform receives the data that vehicles and built-in devices send continuously, processes it and makes it available to applications, such as member apps, breakdown services or analytics for the business.

Why did OpenResearch choose Microsoft Azure for ADAC?

Because the team knew Azure very well. With such a tight timeframe, familiarity with the platform mattered more than any theoretical alternative; it’s what made the pace possible.

What is a lambda architecture?

A data architecture with two paths: a stream path processes events immediately so apps can react in real time. A batch path processes the same data in bulk for analytics and business intelligence. The result is one data foundation for two speeds.

How long has OpenResearch been working on ADAC Smart Connect?

Since July 2021. After sprint 0, the first version of the Smart Connect app was pentest-hardened and production-ready by year end. Since then, OpenResearch has kept developing and running the platform, through trial operation, automated diagnostics and manufacturer data to the roadside assistance pilot in 2026.

How available is the ADAC Smart Connect platform?

Over three years of trial operation, availability was 99.97%. The pilot project required 98.5%.

Does the platform work without a dongle?

Yes. Since 2025, vehicles have also been connected directly via the manufacturer’s interface, for several manufacturers. The app, backend and SCOPE have their own onboarding for manufacturer connections.

Does OpenResearch build such platforms for other mobility providers?

Yes. Since 2016, OpenResearch has built connected-vehicle and IoT systems, from a complete scooter-sharing platform to event-driven processing of device data. We bring that domain knowledge to every new project.