How MOIA is building sustainable urban transportation with AWS
More than half of the global population now lives in cities. With this number constantly growing, cities are looking for ways to introduce more sustainable transportation solutions. MOIA, a young mobility company within the Volkswagen Group, addresses these challenges by providing clean and efficient urban transportation services. They currently operate in the cities of Hamburg and Hannover in Germany. MOIA’s solution -their backend, data processing, analytics, business intelligence workloads- run on AWS. With their ride-sharing service MOIA’s goal is to help cities improve their sustainability through reducing air-pollution and congestion.
This blog post explains how MOIA uses AWS to address the 3 aspects of cloud sustainability, sustainability THROUGH the cloud, IN the cloud and OF the cloud.
Sustainability through the cloud at MOIA
MOIA makes moblity in cities smarter. MOIA’s vehicles do not have fixed routes. Instead, passengers whose start and destination points are similar, share a vehicle that follows a dynamically calculated, optimized route. Sometimes passengers need to walk a block or two to the pick-up spot, as the busses only stop in places agreed with the city authorities.
MOIA uses specially designed modern, comfortable, fully electric, soon-to-be-autonomous vehicles and an app for customers to book their rides. They collaborate with city authorities to provide an affordable, convenient and sustainable transportation service, reduce individual car traffic, and to use road network more efficiently. This is a perfect example of sustainability through the cloud, as MOIA is using AWS technology to solve a broader sustainability challenge.
Sustainability through the cloud touches ecological, economic and social aspects of life. MOIA’s mission to redefine mobility in urban areas, does not only provide an ecological solution that reduces emissions in the cities, but ride-sharing also cuts down transportation cost, and also, less traffic creates nicer communities to live and travel in.
Shared responsibility model for sustainability
Environmental sustainability is a shared responsibility between customers and AWS. AWS is responsible for optimizing the sustainability of the cloud – delivering efficient, shared infrastructure, water stewardship, and sourcing renewable power. Customers are responsible for sustainability in the cloud – optimizing workloads and resource utilization, and minimizing the total resources required to be deployed for your workloads.
Sustainability in the cloud at MOIA
Sustainability in the cloud starts by understanding your environmental impact and then establishing your sustainability goals. These can be reached by increasing resource utilization, eliminating waste in the form of unnecessary compute and data, anticipating and adopting new and more efficient hardware and software offerings, and by using managed services. All of these reduce the downstream impact of your cloud workloads. Just like how MOIA’s vehicle engineers are optimizing bus components and energy use of the vehicles, MOIA cloud specialists make sure their algorithms and cloud usage are optimized with the following best practices:
- MOIA uses serverless services. Friday and Saturday evening, there are highest number of vehicles on the road, except those nights, when there are fewer vehicles operating, the infrastructure scales down automatically to optimize usage of resources.
- Some parts of the workload don’t require real-time data, so this data is batched and processed hourly or daily. This is an example of aligning SLAs that support sustainability goals while meeting your business requirements. MOIA runs bigger analytics jobs at night and on weekends when the energy demand from industry and private consumers is lower. MOIA has optimized their hourly jobs from a run time of 17 minutes to 2 minutes on Amazon EMR cluster by optimizing Spark parameters to enable faster reads of a high volumes of small files.
- Amazon Kinesis buffers the vehicle telemetric data, and the rate limits are set eliminate peaks to ensure smooth consumption of data. This is aligned with the recommendation to optimize workloads for asynchronous and scheduled jobs.
- MOIA adopted AWS Graviton2 instances which are more energy-efficient and offer better price-to-performance ratio compared to other instance families. They regularly update their workloads to use newer and specialized instance types.
- As MOIA is currently operating in Germany, they use AWS Frankfurt region to store and process data, and run applications. This follows the recommendation to store data as close to the consumer as possible, also taking the energy mix of that region into consideration. AWS Frankfurt also leverages Renewable Energy Credits and Guarantees of Origin to cover the non-renewable energy we use in the region.
- The data processing uses spot instances for stateless workloads to increase overall utilization of the cloud and to reduce the sustainability impact of unused resources and save cost.
- MOIA is using Amazon CloudWatch to monitor their workloads. CloudWatch provides metrics to identify underutilized cloud resources. Additionally, by combining cloud and business metrics MOIA defines proxy metrics for sustainability like the amount of computing resources per passenger ride, or the amount of telemetric data stored per kilometer travelled by their bus. They work on long term optimizations of those proxy metrics.
- MOIA uses S3 Storage Lens to identify and remove unused data in Amazon S3 buckets. They also use data lifecycle rules to move data which is not accessed regularly to more energy-efficient S3 Glacier storage.
- Finally, MOIA automates deployments using Apache Airflow. This gives them the possibility to keep their workload up to date, experiment and introduce improvements.
These insights to MOIA’s decisions on architecture and technology choices are examples of sustainability in the cloud.
Sustainability of the cloud
Since the customer is responsible for the sustainability in the cloud, then what is the role of AWS in sustainability for its customers?
AWS takes care of sustainability of the cloud itself. This starts by selecting the location of the data centers. AWS carefully chooses the data center locations to mitigate environmental risk, such as flooding, extreme weather, and seismic activity. Then AWS ensures the sustainable building of the data center. In new constructions, Amazon works with CarbonCure Technologies that injects recycled carbon dioxide into building concrete making floors and walls stronger with less cement. The process also permanently captures recycled carbon dioxide preventing this from burdening the atmosphere.
Next, we need to think about the hardware inside the buildings. AWS takes responsibility for the sourcing of the equipment and directs its vendors to prioritize reuse of generic parts and to maximize recycling of all components that cannot be reused. For some aspects AWS uses in-house designed hardware, like custom-made routers or the purpose-built hardware & software platform AWS Nitro System. This gives AWS the possibility to extend the lifetime of these devices by making them more serviceable and support them longer with driver and firmware updates. Finally, AWS ensures that our waste and decommissioned hardware are responsibly managed to their final endpoint.
Then we need to consider the electricity usage. In 2020 Amazon became the largest corporate buyer of renewable energy. Our ever-growing wind and solar energy projects around the world are contributing to put Amazon (not just AWS) on a path to running on 100% renewable energy by 2025. And if you think of the scale of Amazon, that’s a big goal! To achieve this, every aspect of the operation has to be reviewed and optimized. Let’s look at the example for cooling: When possible, AWS incorporates direct evaporative technology for cooling our data centers, significantly reducing energy and water consumption. Another example of energy saving is using in-rack UPS systems that reduce energy waste from power conversion by about 35% compared to centralized solutions.
Finally, when you choose to use AWS managed services, the teams that build and run services (like Amazon S3) are responsible for optimizing them for power efficiency. With AWS managed services, you pass part of the responsibility to the AWS service teams. You still need to make sure that you don’t perform unnecessary computations or store data that is not needed.
AWS plays a significant role for its customers cloud sustainability goals by laying an environmental-friendly and efficient foundation from buildings and hardware to managed services, optimizing the sustainability in the cloud.
In this post, we explained how MOIA achieves sustainability through and in the cloud, and benefits from the sustainability of the AWS cloud. Visit the Amazon Sustainability website to learn more about the actions that Amazon is taking to ensure sustainable operations and become net-zero carbon by 2040. Refer to the USA, EMEA or APAC reports from Research 451 to understand how AWS data centers are more energy efficient and have lower carbon footprint than a typical on-premises corporate data center. Refer to the Sustainability Pillar of Well-Architected Framework Review for best practices on how to evaluate and optimize your cloud workloads.
We all need to do our part to enable a more sustainable future for the generations to come.