Skip to content

xCog: QA for AI, Robotics, and Autonomy

Quality Assurance for AI, Robotics, and Autonomy

Currently, automated testing for cognitive systems is a relatively new process that requires further exploration. As technology evolves at a rapid pace and gets “smarter”, so do testing methods. Just a few years ago the leap from a facial recognition bot to full out to machine-based surgery would have been considered science fiction. However, these days there’s more science than there’s fiction as the ideas have become reality.

The need for AI, Robots, and Autonomy is growing by the day. While the human population grows, so does the need for technology to accommodate in ways that allow for a higher quality of life. Already, machines play a vital role in the human world, with traffic lights, fraud detection, chatbots, stock trading, agricultural pest control, and the list goes on.

With all of these new avenues and technological advances one question rises to the top, is it safe? Will AI understand human error versus fraud? Can a robot perform surgery and make complex decisions? How can an autonomous vehicle recognize a person crossing the road outside of their programming?

QA Consultants wants to answer these questions and more need by offering a solution. Where to begin? The following are very real issues that are a good place to start a conversation.

Current autonomous robot vehicles quality challenges

 

Limited knowledge

Lack of expertise to properly test an autonomous robot vehicles


Regulatory compliance

Regulatory compliance and government integrations or approval


Long time to market

Once the time to be spent with testing and validating hazard scenarios is high


Physical test high costs

It is not feasible to try reproducing several test scenarios physically


Risk for the robot buyers

No formal guarantee about situations where the vehicle might be not able to avoid obstacles and collide with other objects or people


Limited test coverage

Uncertain about the robotic vehicle’s behaviour in different situations


Late validation in SDLC

Validation and testing are done in the late stages of the robot vehicle development


High complexity

High degrees of complex combinations to be tested to ensure high quality


The Solution

At QA Consultants, we’ve listened to our clients explain their pain points and we’ve created solutions.

  • A proprietary model-based solution that automates the generation of automated test cases for cognitive systems
  • Integration with industry-standard simulation software
  • Automated execution of test cases and result analysis

Accelerating time-to-market with dramatic cost reduction

Benefits & advantages of the xCOG solution

  • Lower cost due to simulation

    Any test scenarios can be performed in a simulated environment

  • Fast time to market

    Hundred of thousands of different test scenarios can be executed quickly

  • Quality & safety assurance

    Potential issues can be identified and fixed before the robotic vehicle goes to production

  • Strategy for a real-world UAT

    Focus on scenarios identified as higher risks to be reproduced in real-word with physical tests

  • Lower risk for the robot buyer

    With a formal quality assurance process, robot buyers can get confidence that it is ready to be interacting in a real environment with people without hazard risks

  • High testing coverage

    The model-based testing approach allows that hundred of thousands of test scenarios are automatically generated and executed

  • Early validation in SDLC

    The software that controls the robotic vehicle can be tested in early stages of development, due to the automated approach and easy validation

Accelerating the automated testing for cognitive systems


  • Faster and economical

    The setup of testing is faster and more economical, once it doesn’t require physical equipment and environment

  • Model-based testing

    Generation and prioritization of test cases are automatically done, using a model-based approach

  • High scalable

    The test is performed in a simulated environment in the cloud

  • Full automation

    Automated tests are automatically generated by the framework

  • Velocity and permutation

    Safety through velocity and permutation: A solution that has automated the creation of automated tests

  • Safer test execution

    No risk to human life or equipment to validating thousands of scenarios that are needed

Market uses

End-to-end Quality

 

Government & regulators

Government, Regulation entities and even insurance companies might be interested in making sure the robot vehicle is safe enough to be running in an environment where people circulate or interact


Robot manufactures

Companies that build a physical robot vehicle. These companies usually are the ones responsible for the entire robot vehicle and the first point-of-contact to buyers or users of those robot vehicles


Robotic systems owners

Companies that buy robot vehicles and have them running in their environment. Examples of Robot Buyers are Airports, Malls, Government, Warehouses, Hospitals, etc.


Funders & investors

Organizations that provide funds and grants for the autonomous robot vehicles ecosystem to invest on technology and solutions to address different needs


AI developers

Software (AI) builders that control and provides autonomy to the robotics vehicles. The intelligence to self-driving and take its own decisions given the scenarios and circumstances it is exposed


How can we get your business to the next level?

Let’s talk. Fill out the form and let us know what you’re interested in doing.