Our client, a global leader in location-based services, was experiencing significant time limitations and a lack of specialized engineers to continually expand the capabilities of their innovative cloud-powered developer platform. Billed as the most reliable road safety technology on the market, our client’s system allows leading OEMs and automotive brands to build advanced geodata-centric solutions and services for autonomous driving.
As our client’s services were quickly gaining traction, demand for new products and features was on the rise. With that came the need for our client to launch more projects and find more specialists to make up for the resulting engineering gap. In addition to rapidly growing development needs, our client faced the necessity to bring on board DevOps and QA engineers to increase efficiency across the entire implementation cycle and deliver the highest quality mapping products.
As a company with long-standing expertise in automotive and location-based technologies, Intellias committed to help our client build an autonomous driving platform and establish DevOps and QA processes.
Our client’s automotive big data solution benefits end users by providing shared real-time and historical sensor data, multilayer high definition (HD) maps, and predictive machine learning. The platform provides an efficient development environment for the rapid creation of proprietary solutions using its shared assets. It also serves as a one-stop marketplace for enterprise-grade, continuously updated spatiotemporal data. With a subscription to the platform’s location services, consumers can easily build and market their custom solutions.
Working alongside our client’s in-house and remote teams, Intellias has been involved in conceptualizing, developing, testing, and deploying the platform’s cloud-based architecture. The entire project is run using the Scaled Agile methodology, ensuring team-wide understanding of performance goals and prioritization of development strategies.
For our client’s Highly Automated Driving (HAD) department, Intellias has set up several teams, each working on a specific part of the project with the cloud software development. At the same time, we’ve recently launched a team for Fully Automated Driving (FAD) whose goal is to build robotaxi technology that will set the stage for the development of software for self-driving cars.
Our engineers are responsible for the full life cycle of 3D HD maps, from eliciting source data to creating and publishing the maps themselves. The map development process is sequential and happens in five major stages:
1. Real-time streaming perception
We develop an Edge Perception stack that allows for HD map observations and crowd-sourced updates using vehicle-mounted sensor systems. The process involves detecting road features in video streams in real time. We then use this dynamic data in the development of self-healing maps. In case of any changes on the roads, maps are updated automatically, and the new map data is delivered to end users in real time.
2. Data collection
To build detailed and credible maps for navigation systems and custom solutions, accurate information on road attributes needs to be collected from around the globe. Our teams gather data from mobile cameras, sensors, and GPS devices on the road to locate traffic lights, signs, poles, stop lines, lane markings, roadside barriers, junctions, etc. We’re also developing an Android application that helps to optimize data collection processes across the world.
3. Data aggregation, processing, and filtering
Our engineers handle massive datasets from a variety of sources, including from core maps and on-vehicle sensor systems. All road image data is aggregated, processed, and filtered via a streaming file system for further validation and intelligent analysis using machine learning models. All data can be consumed in an NDS (Navigation Data Standard) format to avoid vendor lock-in and ensure interoperability across systems.
4. 3D map creation and maintenance
We compile and update live 3D maps that enable precise positioning for lateral and longitudinal control of vehicles. These multi-layer maps contain details at several levels (roads, lanes, lane groups and individual lanes, geometry) and are constantly enriched with incoming information on new road attributes such as signs, markers, crosswalks, bicycle lanes, and objects.
5. Map delivery
Data-intensive HD maps are generated and delivered to customers online as a geographically tiled and functionally layered data service suitable for direct-to-vehicle and OEM cloud consumption. These maps are also an indispensable source of data for our client’s many in-vehicle software development programs.
In addition to helping our client build their map data platform, our experts hold full responsibility for DevOps and QA processes on the project. When our cooperation just started, many manual workflows within our client’s legacy system needed to be automated to avoid extra overhead and multiple errors. Our client was also looking to build QA capacity that was lacking at the time. Intellias engineers have established DevOps and QA Centers of Excellence for our client, providing full-scale consulting and services to automate and test our client’s system.
We’ve initiated and launched a major migration of our client’s solution to a new, cloud-based, scalable system. One of our recent focuses has been on the transition from Gerrit to GitLab repository management, which will allow our engineers to set up pipelines, view updates, and manage code more efficiently and in one place. We’re also in charge of several other migrations that set the foundation for a more productive cloud application development platform.
Together with our client’s development teams, we’ve successfully completed two major releases, creating digital platform design and empowering customers’ systems with fresh data and a pack of new features to optimize navigation. The most significant ingredient in our success is our engineering know-how in computational geometry, graph theory, unsupervised and online machine learning, image processing, and predictive models. For our client, this cooperation has proved an exceptional opportunity to speed up time to market, meet their business commitments to customers, and mitigate delivery risks.
Our contribution to data platform development for autonomous driving brings top technology for assisted and autonomous driving to the automotive market. The HD mapping platform we built opens up the possibility for enriching multilayer maps with real-time data that can be exchanged among connected vehicles. Our engineers’ expertise and quality improvement initiatives have provided the driving force to achieve ultimate data accuracy and road safety.
The solution we’re developing brings these benefits to customers:
- Global map coverage and near-real-time map updates
- Multi-layer maps that can resolve up to 90–95% of mapping data errors
- Access to rich map data layers, including classic 3D views, road-level and lane-level details, and an extended list of attributes
- Size-optimized maps due to compact data storage formats
- Access to built-in services such as routing, geocoding, and rendering
- An architecture that lets developers build complex applications
- A managed environment for secure data deployment
Our partnership with Intellias allows us to free up our in-house resources and focus on other strategic projects while managing business requirements for the product Intellias has been delivering. Their team did a great job by optimizing the development, DevOps, and quality assurance processes, resulting in increased efficiency and a 40% reduction in costs. Intellias impressed us with their solid talent acquisition process and quality standards.