Advanced Computing in the Age of AI | Saturday, January 22, 2022

AWS Dives Deeper into Enterprise Robotics at re:Invent 2021 

As Amazon Web Services (AWS) opened its re:Invent 2021 technology conference in Las Vegas this week, the news was not only about cloud computing. Also getting top billing are the company’s strategic moves in advancing its offerings in the growing world of enterprise robotics.

To help companies and robotics developers build and deploy applications that enable fleets of robots to work together smoothly, Amazon launched a public preview of AWS IoT RoboRunner, a new robotics service that connects robots and work management systems so they can be managed through a single system view.

Also unveiled at re:Invent 2021 was an AWS Robotics Startup Accelerator, a four-week technical, business and mentorship accelerator program that will be open to startups around the world. The accelerator, which is being organized and offered in collaboration with non-profit robotics innovation hub, MassRobotics, aims to provide guidance and mentorship for robotics startups to help prepare them for greater success in the marketplace.

AWS IoT RoboRunner

Created to help developers and enterprises build applications to manage fleets of robots, the new AWS IoT RoboRunner robotics services follow the company’s 2018 launch of AWS RoboMaker, which is a cloud-based simulation service that enables robotics developers to run, scale and automate simulation without managing any infrastructure.

Eric Anderson of AWS

AWS IoT RoboRunner builds on the RoboMaker capabilities by helping enterprises and developers make their robots easier to connect. RoboRunner uses the same capabilities that Amazon uses in its own product fulfillment centers when orders come in, Eric Anderson, the general manager of AWS Robotics, told EnterpriseAI.

“RoboRunner started as a program within our internal Amazon robotics teams,” said Anderson. As the team was scaling up Amazon's robot fleets with new types of robots in new configurations, they discovered three common problems which kept appearing – figuring out how to connect robots to the edge and into the cloud; finding new ways to interoperate the robots as new hardware configurations, new products from multiple vendors and distinct project needs appeared; and how to optimize workflows and orchestrate work among those varied robots.

“So, we built RoboRunner as an internal application to drive and power Amazon's fulfillment and logistics robotics,” said Anderson. Quickly, AWS started hearing questions and concerns from its own customers including agricultural products vendor AGCO and global system integrators like Accenture, which were already working to implement their own robotics infrastructures.

“They were finding the same problems as well,” he said. “So, earlier this year we decided to build on top of this internal application that we used to orchestrate and manage robotics for Amazon's fleets, and we created AWS IoT RoboRunner. We have more than 350,000 robots deployed in our environments today, ranging from logistics to fulfillment, both the structured field [robots] from Amazon Robotics, formerly Kiva Systems, but also other third-party and customized robotics products”

From its birth and use inside Amazon to get customer orders picked, packed and shipped, RoboRunner was then adopted by enterprises to assist in manufacturing, logistics and other use cases, said Anderson.

AWS IoT RoboRunner is now available in public preview to customers in the U.S. East (Northern Virginia) and European (Frankfurt) regions, according to AWS.

AWS Robotics Startup Accelerator

The AWS Robotics Startup Accelerator will be offered virtually in the first quarter of 2022 to robotics startups that want to adopt and use AWS to boost their robotics development, according to organizers. Participants will also receive support from industry and AWS experts to help rapidly scale their businesses.

Registration for the accelerator is now open through Jan. 16, 2022. MassRobotics will provide expertise involving working with venture capital and startup-scaling to help AWS drive a select cohort of robotics startups through an intensive, advanced curriculum specially designed to accelerate their growth using the AWS Robotics services.

Participation is open to privately-held startups in early stages, with less than $10 million in revenue and less than $100 million raised, involving robotics technology or related areas. Startups do not need to currently be using AWS to apply.

A selection committee from MassRobotics and AWS will evaluate applicants based on a wide range of factors, including the uniqueness of their technologies, the ability to roll out products within 12 months of completing the program, clear and scalable business plans and market focus as well as startups with more than six employees.

Once the four-week accelerator program is completed, the selected startups may continue to have engagements with AWS and MassRobotics experts and mentors and may get opportunities to tell their customer stories through AWS marketing channels.

The accelerator program is getting support from AWS because the field of robotics is expanding quickly and new innovations will help the industry as well as customers, said AWS’ Anderson.

“There are a number of startups that are out there trying to innovate the next generation of robotic equipment or products, but they need help to adapt cloud architectures to their robotic applications and make them work for their customers,” he said. “So, we partnered with MassRobotics to create this accelerator program, which really combines technical training in terms of network integrations and software applications and how to leverage the power of AWS cloud for robotic application development. [The program] will provide training on things like fundraising, incorporation and how to help small startups understand how to scale their businesses to the next level.”

These are some of the custom robots used by Amazon for order fulfillment in its warehouses and facilities. Image courtesy of Amazon.

 

Add a Comment

EnterpriseAI