How we created the Embedded Logger to help customers collect statistical data for their cars

This software solution will be used in more than 1 million cars every year

In April 2018, Objective Software GmbH, a company with more than 20 years in the automotive and financial sectors, offices in four countries, and more than 130 employees, turned to us as an embedded software development company to get help for a software project. At the time, they were looking for a perfectly coordinated dream team of experts in Autosar software architecture and embedded software development with customer processes experience.

What the project was about

The goal was to create an embedded logger for collecting statistical data on cars, suitable for big and small ECUs, supporting both classic and adaptive platforms.
One day, automobile manufacturers will remotely update software systems and monitor engine conditions or powertrain performance. Big Data will help in fleet management, in planning roadways and traffic flows, in creating customized insurance products, etc.
Collection and analysis of diagnostic data in the automotive industry is of paramount importance for both lifecycle support as well as tailoring product development.
The embedded logger can provide any type of statistical data from a car, from engine or exhaust system performance to the frequency of using car wipers. For automobile manufacturers, this software solution helps them gather the most complete and accurate data during new technology development or improvement, as well as in the aftermarket maintenance, process optimization, therefore saving customer's time and money.
Similar Cases

Technical side of the project

Tools:
  • Visual Studio 15, Autosar Builder 2019, git/gerrit, Quality tools: QAC, daVinci Vector, Adaptive Demonstrator, gTest, cUnit
  • Technologies: C90/C++14
  • Protocols: SOME-IP, CAN-TP, UDP
  • Used Autosar Standard: classic 4.3 and adaptive 18-03

The process of embedded logger creation

The team that was working on this project consisted of the software architect, who was a team leader based in Munich, and three embedded software developers, who were participating in the project remotely.

Such an organization of work made it possible to make decisions on the spot and at the same time save client resources, thanks to remote collaboration.

We used an agile approach, with sprints of two weeks and product increments of four weeks. Each PI was concluded with the formal delivery process.

A well-performed requirements analysis is a key element for project planning and resources estimation. That’s why, before starting to write code, we thoroughly analyzed the requirements in their entire scope and worked out the architecture and software detailed design. All technical requirements were divided into independent tasks; for each of them, a ticket (or tickets) with story point estimation was created. Based on this estimate, a development plan with final product delivery at the end of August (for classic) and the end of November (for Adaptive platform) was created. This plan matched well with the customer’s expectations.

  • To enable early integration, we started with creating the interfaces, followed by the functional code. This practice has proved itself well and we used it for each next product increment.
  • We have faced minor difficulties due to some flaws in the requirements. These were resolved mainly beforehand, as we have organized fast iterations with the customer requirements owner, thus usually having ready solutions before planning the next product increment.
  • Some of the interfaces that we needed to connect with were not ready on time. We had to think them through ourselves. Later, the customer created the missing interfaces exactly as we had designed.
  • Another problem during development was connected to system tests. As we were responsible for the unit and module tests only, we could provide verification for this small scope. But the customer’s department had a quite extensive time lag in test readiness, such that the system tests were carried out after the final delivery, and until that time we did not have full verification of the system scope. The sole countermeasure we could rapidly propose was intensive code peer reviews with thorough functional checks. This has helped us a lot, but some bug fixing iterations were nevertheless carried out afterwards.

The result of 8 months of work

1
The first phase of the project was successfully completed in December 2018. All issues we had during development were solved successfully.
2
The software turned out to be very RAM-consuming. This was a result of the functional requirements proposed. We have provided different options for customization, so that RAM consumption can be reduced. The functional requirements were optimized afterwards to make the product more compact, and we have implemented more sophisticated memory management and other optimizing solutions in the next few months
3
The embedded logger for collecting statistical data will be used in more than 1 million cars every year.
“One day, automobile manufacturers will remotely update software systems and monitor engine conditions or powertrain performance. Big Data will help in fleet management, in planning roadways and traffic flows, in creating customized insurance products, etc."

Here is what Objective Software GmbH says about our cooperation...

Amanda Peterson Marketing Director at Junk Food Co.

“We started to work with this Company 4 years ago. We had only positive experience and implemented a lot of projects together. We’re totally satisfied and now we know that we can rely on this Company experts any time. Now our clients receive their products faster so that they are more satisfied with our services. They deserve the highest grade!"

October 28, 2017

Let's start developing your embedded software project together!

Complete the form below and we'll contact you to answer any your question.