This is our latest post regarding our software ANDI (Automated Nuclear Damage Inspection) and incorporating identification capability with 360 video data.
The 360 camera sector has advanced significantly over the past few years, camera resolution and image quality has improved greatly as well as advancements in software processing to provide multiple different ways of viewing the captured images and video. Cerberus Nuclear has been keeping up to date with latest developments with an aim of using this technology in the nuclear sector.
Cerberus recently developed ANDI (Automated Nuclear Damage Inspection) for Sellafield Ltd. The software automatically identifies key areas of damage from inspection videos and is currently being used by Sellafield to accelerate damage inspection tasks. The software is built into a user-friendly interface and supports the creation of reports and logging of key identified features.
Building upon our knowledge we are currently testing the use of 360 camera and video data with our custom computer vision algorithms, including ANDI. Some key advantages of using 360 data for automated damage inspection is that the orientation of the inspection camera is no longer a factor as images capture the full 360 degrees.
Similar technology is currently being used in autonomous vehicles for object identification and distance determination.
Footage obtained from a Cerberus Nuclear test car.
Our preliminary testing has proved to be very successful and we have overcome some of the challenges inherent in working directly with 360 data sets. The prototype software we have created demonstrates the capability of combining both 360 image technology with our bespoke computer vision algorithms.
Our goal is to continue the development of ANDI so this highly useful and innovative technology can be put to good use solving a wide range of challenges in the nuclear sector and beyond.
Look out for future updates, if you would like to learn more don't hesitate to get in touch at nuclear@cerberusnuclear.com.
Sellafield LINC: Image Processing for Assessing Package Integrity
The Challenge
LINC with Sellafield Ltd is a scheme that encourages SMEs at local and national level to collaborate and deliver innovative solutions to support the mission at Sellafield. LINC challenge 42 was titled ‘Image Processing for Assessing Package Integrity using Machine Learning’ and set the challenge as follows:
“The long-term storage of nuclear waste is at the heart of the nuclear industry in the UK. As a requirement this material must be examined on a regular basis, which generates a vast amount of data to be reviewed. Presently this is done manually and takes a lot of time and is vulnerable to human error."
Cerberus Nuclear’s Data Science team proposed the development of software containing a trained machine learning computer vision model that would be capable of automatically recognising issues that could affect integrity of the package. Our team use the ‘Agile’ design methodology, which incorporates software testing in short, focused development cycles; ideal for the project.
We were delighted when Sellafield Ltd chose our solution ahead of some tough competition.
ANDI: Automated Nuclear Damage Inspection
ANDI is a high-quality user-friendly software program that utilises computer vision machine learning for automated identification of damage from externally supplied video.
The software allows detailed examination of product can inspection videos, automatically identifying damage such as scratches, dents and corrosion. The neural network within ANDI uses a cutting-edge R-CNN approach for image analysis and was trained using previous examples of damage.
Damage identified is highlighted within an embedded video player, which allows users to quickly skip to areas of interest and examine results frame by frame to inspect the exact moment(s) that damage has been detected. The confidence level of identified damage can be customised by the user with damage highlights switched on or off to assist with detailed inspection.
An integrated inspection report system was incorporated into the software to allow users to make notes and log frames for easy follow up review. The software also allows the processing of multiple batches of inspection videos with minimal user interaction. This allows the review of multiple processed results within a single session.
The algorithm for the Sellafield challenge uses an extension of the R-CNN called the Mask R-CNN. The R-CNN algorithm (Region based Convolutional Neural Network) can detect and classify objects within images, it focuses on variations of colour, texture and scale within an image to form a region.
“I really like the look and feel of the software and I’m impressed how well the neural network is identifying the key elements of damage, it’s very good!”
“We are delighted that Cerberus Nuclear helped make a difference at Sellafield Ltd. The Data Science team have delivered a great solution, bringing modern techniques to the nuclear industry.”
Daniel Cork, Director, Cerberus Nuclear
Sellafield Ltd are currently using ANDI to enhance their damage inspection workflow which, prior to using the software, had taken many man-hours to identify, categorise and log.
Cerberus Nuclear are proud to announce ANDI has recently been a key feature for inTechBrew. inTechBrew promotes the latest high Technology Readiness Level (TRL) nuclear industry innovations across UK and Europe.
Machine Learning Computer Vision
Previously, the team developed a custom computer vision object identification algorithm to identify cars and lorries on a motorway using R-CNN object detection method.
The algorithm identified the number of objects, object type (car, lorry, etc.), object colour, object speed and confidence in match. In addition, a report was generated that summarised the information gathered over time. The development was to demonstrate validity of use for stopped car identification as well as traffic monitoring purposes.
The Sellafield Ltd LINC challenge aligned well with the previous development work already performed and paved the way for the creation of ANDI. Additional technical challenges such as variable lighting, frame blur and reflections had to be overcome as well as creating a custom user-friendly interface that met with Sellafield Ltd requirements. The processing time for the computer vision algorithm was also enhanced.
Cerberus Receives ARC Funding to Develop Criticality Safety VR Training Software
Cerberus Nuclear is a hub for innovation in criticality safety and radiation shielding and we are pleased to announce that we have made a successful application for funding from the UK’s Alpha Resilience and Capability (ARC) programme. ARC was created by BEIS (Department for Business, Energy & Industrial Strategy) to ensure that the UK retains its world-leading alpha capabilities: from operations and maintenance, to high-end R&D and design. This cross-industry consortium includes the National Nuclear Laboratory, Sellafield Limited, AWE and ONR.
Over the last two years Cerberus Nuclear has developed CARTA, a concept for criticality safety VR (Virtual Reality) training software, which we successfully presented at ICNC2019. Uniquely, CARTA uses a machine learning algorithm to predict k-effective 'on the fly' for a given system, such as an alpha facility glovebox. When coupled to a VR headset, CARTA gives users an immersive experience of the facility environment and the effect of their actions on the system’s reactivity.
The ARC funding will support the next phase of development, to refine the concept into a software package for members of the ARC consortium to use. CARTA will deliver tangible benefits directly to operators on plant, criticality safety specialists and other stakeholders in criticality safety. The software package will use a variety of scenarios in desktop and VR environments, to provide intuitive user interfaces. The underpinning data will be based on accurate modelling of the neutron physics, providing a realistic environment for trainees to improve their understanding of the complexities associated with criticality safety.
The specifics of the training scenarios will be guided by a Technical Steering Committee, comprising stakeholders from the various ARC member organisations. This will ensure that the training scenarios are relevant and can be effectively integrated into their existing training programmes.
We are now actively seeking organisations that would benefit from bespoke criticality safety training scenarios. If you would like to discuss your idea, please get in touch using nuclear@cerberusnuclear.com.
Eddy - MCNP & SCALE Html Generator
Cerberus Nuclear has created Eddy, an open-source Html output generator for MCNP and SCALE. The function of Eddy is to parse MCNP and SCALE output files into an easy-to-read and user-friendly format. Eddy has been written to work for both radiation transport and criticality calculations.
Eddy collates key information from an output file so that it can be quickly reviewed. Normalisation factors can also be specified to simplify interpretation of tally outputs.
Eddy Html outputs include:
Normalised Tally Results with Error
Highlighted Statistical Checks
K-effective and Error
Comments and Warnings
Cell Mass and Volumes
Particle Populations
Full MCNP Input
Eddy is simple to use from the command line or via its built-in interface. Hyperlinks within the Html enable the user to navigate to the required part of the output with ease.
The Html output from Eddy assists in the preparation of technical reports supports QA processes and improves workflow efficiency. The contained nature of the Html output and its small file size also facilitates the sharing of calculation outputs for independent review purposes.
If you would like to provide any feedback or would like to request additional features please get in touch by emailing nuclear@cerberusnuclear.com
Object Recognition - Computer Vision
We have developed a custom computer vision object identification algorithm. As a demonstration, we trained a neural network to identify cars and lorries on a motorway using R-CNN object detection.
The algorithm identifies a number of objects, object type (car/lorry etc.), object colour, object speed and confidence in match. In addition, a report is generated that summarises the information gathered over time.
Cerberus Nuclear has access to the necessary computing power for near real-time, high-accuracy identification. The example was produced to demonstrate validity for stopped car identification as well as traffic monitoring.
In addition to this example, computer vision has a number of nuclear applications that we will be looking to build upon in the coming weeks.
If you would like to know more or if you have a specific challenge, which you think computer vision may solve, don't hesitate to get in touch.
