Developing remarkable technology.
The NDE group is an innovation based group of companies, where we continuously invest in pioneering innovation and custom-built solutions for the challenges of today and tomorrow. We actively collaborate with our customers, supply chain and universities to explore the application of new technologies for the benefit of our customers. Born from a need for non-destructive testing and high-end inspections to avoid asset failure, we have evolved our research and product development into signal processing and robotics.
Our neurological inspired signal enhancement processing capability has applications beyond the non-destructive testing domain, into many defence related applications involving electro-optics, sonar, acoustics, IR and more. We are passionate about innovation and the application of emerging technology to a range of applications. Our research and product development group is focused on applying new and innovative technology to a broad range of defence and non-military applications, covering scanning, signal processing, data management and analysis.
NDE technology is an innovation led company researching and developing new technologies and their applications in the NDT market. Our work encompasses concept exploration, design, prototyping and small scale manufacture of items for usage in the NDT services market. We actively collaborate with partners including Universities, customers, our supply chain and NDT technicians in pioneering innovation and custom-built solutions for the challenges of today and tomorrow.
We currently work in the following product areas:
- Control systems
- Acquisition systems
- Analysis systems
- Hand held scanners
- Robotic controlled scanners
What is TrueSignalTM?
TrueSignalTM is a biologically inspired neurological model signal processing application. TrueSignalTM processing capability has applications in a range of sensor domains and applications. Our software provides the processing and analysis of raw sensor data to enhance signal to noise ratio, data compression and sub-pixel target detection. We actively seek to work with sensor and application partners to enhance the effectiveness of applications such as target detection, target tracking, AI/ML classification and object avoidance.
The deployment models for TrueSignalTM include:
- As a stand-alone software application with domain specific graphical user interface and functionality.
- As a dynamic linked library enabling integration into existing sensors and applications to enhance the effectiveness and overall outcomes.
- An FPGA implementation for hardware level integration into embedded systems.
Due to it’s neurologically-inspired architecture, TrueSignalTM is self-adapting across a wide variety of sensing technologies to provide domain agnostic capability.
The following key discriminators which have been demonstrated during our research development activities when applying TrueSignalTM to recorded data:
- Sensor agnostic processing covering a wide range of sensors including EO, IR, X-ray, acoustics, sonar and video.
- Multi-sensor detection and tracking of small and low-signature threats in cluttered environments.
- Enhancing existing microphone sensors for acoustic detection of unmanned aerial systems in highly noisy environments.
An example of TrueSignalTM image enhancement processing applied to photographic imagery.
TrueSignalTM has a vast array of demonstrated applications areas; from low-level signal-filtering to task-driven feature enhancement.
We actively seek to work with sensor and application partners to enhance the effectiveness of applications such as surveillance, specialist sensing, automated detection, and remote awareness.
In the awareness and detection tasks TrueSignalTM can provide adaptive noise suppression to existing sensor systems, improving tracking range and confidence. In visual detection tasks, TrueSignalTM can improve detection confidence by up to 60 times in cluttered environments.
As an image post-processing tool TrueSignalTM enables one-touch filtering for vastly improved visual acuity of weak defects and features, reducing analysis and diagnostic time.
In trials with off the shelf AI/ML systems, the enhanced data from video recordings was found to have dramatic impact on the AI/ML capability to detect and classify different entities in a street scene, improving detection rates by an average of 16.2% and enabling earlier detection in 74% object instances with no impact on the classifier confidence.
In the domain of acoustic target detection of UAV’s, it has demonstrated capability in improving automatic detection range by 33% and PSNR (Peak Signal to Noise Ratio) by 10dB.