CAVs (Connected and autonomous vehicles) can change the way people work and travel in cities. During the COVID-19 pandemic, there has been a significant increase in the use of personal vehicles leading to altered travel behaviour and expectations. During the pandemic everyday restrictions have increased the demand for driverless transportation, deliveries and contactless operations, driving the need for autonomous driving technologies and adoption of connected/ autonomous vehicles. And while the long-term impacts of the COVID-19 pandemic on the travel and transportation industry is still evolving, new possibilities for the autonomous vehicles industry and intelligent solutions hold tremendous potential for addressing the challenges that the industry is facing.
Connected and autonomous vehicles also play a crucial role in supporting pandemic mitigation actions by easing the transportation of necessary medical supplies and food to infected areas while reducing potential exposure of drivers.
Connected and Autonomous Vehicle Development
Developing optimal solutions for connected and autonomous vehicle systems requires the collection of massive amounts of data (generated by the vehicle’s sensors and cameras) with the capability to store and manage the data with high-performance computing capacity, advance deep learning, and real-time processing.
Cloud based solutions are now highly mature and can easily provide the backbone to build fully functional connected and autonomous vehicle systems, while taking away the burden of managing the infrastructure. And AWS plays a vital role by providing a full line of services to support the system development and deployment.
AWS’ highly scalable, virtually unlimited compute service and storage, coupled with advanced deep learning frameworks helps collect, ingest, store, and analyze connected and autonomous vehicle data to support the development of autonomous vehicle technology.
Source – https://aws.amazon.com/automotive/autonomous-driving/
Source – https://aws.amazon.com/automotive/connected-vehicles
Reference Architecture for Autonomous Vehicle System
- Ingest: AWS IoT Core, AWS IoT Greengrass, and Amazon Kinesis data firehose collect and transmits near real-time vehicle data.
- Identify low-quality data and remove or transform low-quality data.
- Improve the quality of data by GPS location combined with weather conditions. Then, synchronize the data.
- Detect Scene: Amazon EMR, Amazon DynomoDB, Amazon Elasticsearch with 3rd party tools (identifying the vehicle performing a lane change, etc.)
- Data lineage: With the help of Amazon Neptune and Glue Data Catalog.
- AWS Rekognition and Lambda help to Identify and blur faces. Amazon Rekognition Provides video and image analysis to your applications.
- Amazon SageMaker Ground Truth or third-party labeling tools Perform automated labeling on raw data/anonymized data.
- AWS AppSync, Amazon QuickSight (KPI reporting and monitoring), or another tooling for visualization will provide an advanced analytics and visualization toolchain with search function for particular situations.
References:
https://aws.amazon.com/automotive/autonomous-driving/
https://aws.amazon.com/solutions/implementations/aws-connected-vehicle-solution/#
https://aws.amazon.com/rekognition