AWS Robotics Services

AWS Robotics services

Robotic services have come a long way to become part of our day-to-day activities. For example, robotic services, or bots as we call them help us do daily internet activities like paraphrasing for an assignment, replying to emails, prototyping designs, and so many others. Amazon has taken a step further to make things easier for humans by dabbling into robotic services through their well-known web service called AWS.

AWS RoboMaker is a cloud-based simulation service that makes it possible for developers of robotics to run, scale, and automate simulations without having to manage any infrastructure.

Use cases of AWS Robotic services. (Robomaker)

• Automated Regression Testing: Automate testing within a continuous integration and continuous delivery (CI/CD) pipeline.
• Training of Reinforcement learning Models: Utilize a large number of iterative trials to train reinforcement learning models.
• Multiple Robot Testing: For testing, connect multiple simultaneous simulations to your fleet management software.

1. Simulation

To build and test applications for autonomous robotics can sometimes become challenging, and complicated, and take up a lot of resources. On a normal note, developers write codes that attend to a wide array of software solution deployment scenarios. These codes are then integrated and the application scenarios are tested on robotics hardware in physical environments.

 The manual method of this process is actually stressful, time-consuming, and also takes up a lot of resources (requires very expensive hardware). The update cycle of the manual process is also very slow and can not be easily scaled.

When the simulation is used, developers have easy access to test applications in virtual environments which enables them to increase testing coverage, accelerate development velocity, and curb coding errors to a substantial level.

One of the most prominent reasons simulation is used is its ability to test and train reinforcement learning and synthetic data generation. Scaling simulations unlocks certain features or abilities like conducting robust regression testing, reinforcement learning, and synthetic data generation.

Simulation with AWS Robotic Service (Robomaker).

AWS RoboMaker is a service that is fully managed and it allows developers, QA, and DevOps engineers to seamlessly create simulations and run simulations without provisioning or managing any infrastructure. Robomaker has made simulation at scale, affordable, and very accessible to so many robotic companies. Robomaker was able to achieve this by providing tools that developers can use to test and iterate codes in 3D environs.

AWS Robomaker supports large-scale as well as parallel simulations, and scales based on the complexity of the scenarios which are being tested.

With Robomaker simulation service, companies that deal in robotics can make machine learning and robotics application testing easier, faster, cost-effective, and more robust.

• AWS runs large-scale, parallel simulators

With the Robomaker batch simulation API, developers can launch a large batch of simulations with a single API call seamlessly. Developers can increase the complexity, scale, and frequency of their tests by running large-scale simulations through an API, which speeds up the development lifecycle and makes testing more robust.

• Flexibility of Simulators

RoboMaker Simulation can be used to run the robot software and simulator of choice, including ROS, custom robot applications, Gazebo, Unity, Unreal, and Nvidia Isaac-based simulations. You can run your simulations in the cloud with RoboMaker without having to manage, configure, or provision any infrastructure.

• WorldForge

Robomaker WorldForge gives an avenue for the possibility of users being able to create hundreds of pre-defined, randomized simulation worlds that imitate real-world scenarios without engineering investments or managing world generation infrastructure. Without a simulation world generator, the development of a simulation is expensive, takes time, and requires the skill of 3d modeling and knowledge of simulation engines.

WorldForge makes it simple to increase the scale, speed, and variance of simulation, which improves production code quality and speeds up time to market. WorldForge now provides indoor home environments with configurable floor plans and furnishings.

• Sample Applications: Sample robotics applications are included in AWS RoboMaker to help you get started quickly. Voice command, recognition, fleet management, and monitoring capabilities are typically required for intelligent robotics applications, and these serve as the foundation. The simulation application code and the robotics application code, which provide instructions for your robot’s functionality, are included in the sample applications.
• ROS Extensions: Services like Amazon Rekognition for image and video analysis, Amazon Lex for speech recognition, Amazon Polly for speech generation, and Amazon CloudWatch for logging and monitoring are among the RoboMaker cloud extensions for ROS. Each of these cloud services is available as open-source ROS packages from RoboMaker, allowing you to use cloud APIs to extend your robot’s capabilities within a recognizable software framework.

• Hardware Developer kits: Adding advanced functionality to robots and building them requires developers to make a lot of choices. AWS partners have developed a number of robotic development kits that include comprehensive hardware and software solutions, pre-installed software, extensive documentation, and tutorials to speed development and eliminate uncertainty.

Some of these hardware developer kits are listed below:

• Intel – UP Squared RoboMaker Developer Kit: This is an easy way to begin a robotic project as its starter package is designed to make using AI in the user’s robotic project, seamless. The kits also provide tutorials on how to build hardware from model level and also how to use cloud service to shorten the development time.
• Nvidia – JetBot AI Kit Featuring ROS & AWS RoboMaker: Nvidia speeds up robotic advancement from Cloud to Edge with AWS RoboMaker. Robotic simulation and development can now be effortlessly finished in the cloud and sent across a huge number of robots and other independent machines controlled by Jetson.
• Qualcomm – Robotics RB3 Platform with integrated support for AWS RoboMaker: Support for Amazon Web Service’s AWS RoboMaker by Qualcomm Technologies is transforming robotics innovation. The Qualcomm Robotics RB3 platform gives developers the tools to build robots that can accelerate innovation, revolutionize logistics, and improve our day-to-day lives. These tools include high-performance heterogeneous computing, on-device machine learning and computer vision, high-fidelity sensor processing for perception, odometry for localization, mapping, and navigation, and connectivity to 4G LTE and Wi-Fi.

For the pricing for simulation according to amazon, you only pay for what you use.

The pricing per usage is as follows:

• 1 Simulation Unit, or SU (1 vCPU and 2 GB of memory) – $0.40 per hour
• 1 GPU Unit, or GU (1 GPU device) –  $1.50 per hour
• World Generation – $1.50 per world
• World Export – $5.00 per world

The AWS RoboMaker advancement climate runs on AWS Cloud9. The RoboMaker development environment can be used at no additional cost in addition to the standard AWS Cloud9 prices.

Let’s consider a user that created his RoboMaker development environment on an m4.large Amazon EC2 instance with 8 GB of EBS General Purpose SSD (gp2) volume attached. Consider that the user also develops his robotics application in this environment eight hours per day and 20 days per month.

The pricing will be as follows;

Total instance hours used: 8 hours/day * 20 days/month = 160 hours

Amazon EC2 price per hour of on-demand m4.large instance: $0.10

Amazon EC2 charges: 160 hours * $0.10/hour = $16

Total GB-month of EBS volume used: 8 GB * 1 month = 8 GB-month

Amazon EBS price per GB-month of provisioned gp2 volume: $0.10

Amazon EBS charges: 8 GB-month * $0.10/GB-month = $0.80

Total charges: $16 + $0.80 = $16.80

Robotics is the future of learning, working, and even our daily chores at home. In the present, as we speak, we have been able to harness only a little from the advantages of robotics and it has led us to live a very stress-free life. Robomaker is helping developers look into robotics more for a smoother operation of our daily activities and even the complex matters we don’t have a total solution for, like cancer. With robotics on the rise, in the nearest future, humans will be able to live a totally stress-free life.