“AWS Is How”: iRobot Increases Testing Speed for Better Customer Experience

iRobot uses a simulation service on AWS to test new code using different scenarios and robot types. Tests can be run in parallel, and they run faster than in real time, which means that the company can test a software build across its entire product line.

Hotel Vacation

In 1990, iRobot was founded with the mission of making practical robots a reality. After over a decade of producing robots, the company introduced us to the Roomba robot vacuum, a product that has revolutionized the way we clean our floors. Since then, iRobot has sold more than 40 million home robots worldwide. For each new release of these products that make our lives simpler, iRobot spends several months testing the hardware and software to confirm that the robots perform optimally in any scenario. 

As home-cleaning robots like the Roomba robot vacuum have steadily grown in popularity, so too have consumer expectations of such cleaning robots. To meet this customer demand and to help clean homes even better, iRobot has continued to roll out innovations like self-emptying bins, Smart Mapping to help robots learn home layouts, and voice-activated intelligence powered by Amazon Alexa.

For iRobot to deliver this new technology, much of which is software driven, it became more important than ever that the company increase its testing capacity. This would help the company more quickly identify potential issues and push updates to its products faster. 
 
iRobot decided to move its simulation testing environment to Amazon Web Services (AWS) because it already relied on AWS solutions to monitor its robot fleet. Previously, the company could run about 1,000 tests per day on real robots, but now, iRobot can run more than 20,000 tests per day in simulation. 
 
This improvement means the company can validate test scenarios in hours instead of weeks, helping it to find code quality issues that it wouldn’t have been able to catch before. And catching issues in development instead of production creates a higher quality baseline. This leads to higher pass rates and decreased software time to market, meaning that customers get more frequent updates and improvements to the products that they rely on. 
 
Creating a testing environment in the cloud, where software and services run on a global network of servers instead of on local computers, gives iRobot an advantage in developing its software. The company uncovers potential problems with code using automated regression testing, a software testing technique that uses computer-based tools to test software after it’s been changed or updated.

iRobot uses a simulation service on AWS to test new code using different scenarios and robot types. Tests can be run in parallel, and they run faster than in real time, which means that the company can test a software build across its entire product line. iRobot has completed tests using hundreds of scenarios in home-testing labs to make sure the robots can adapt to different home layouts and environments. Using the simulation service makes it possible for the company to run 20 times the number of tests that it could using real robots, giving iRobot a better measure of hardware and software stability. The company can find problems faster, produce higher-quality code, and ultimately provide a better experience for customers.

Mercure Hotel

Behind the Innovation. AWS is How.