IDRD: In-Line Diabetic Retinopathy Detection

Challenge

Diabetic Retinopathy (DR) is the leading cause of vision loss in adults aged between 25 and 60, with a recent study estimating that the number of people worldwide with DR will grow to 191 million by 2030. Elsewhere, it has been estimated that the cost of DR to the UK economy will reach £97,000,000 by 2036. When DR is caught early, treatment is effective at reducing or preventing vision loss and while a very small number of countries have national screening programmes for DR detection in place, they are not commonplace and are costly to run.

Solution

In-line Diabetic Retinopathy (IDRD) is a light weight Computer Vision and Artificial Intelligence solution for detecting the presence of DR indicators in retinal images on-the-edge, in the specialist fundus cameras that take these images. The solution has been developed around the Intel Movidius Myriad2 Vision Processing Unit (VPU), which pre-processes the retinal images using its on-board image filtering capabilities and then applies an Artificial Intelligence Convolutional Neural Network (CNN) to assess the retinal image for DR indicators. The solution has been designed to assist ophthalmologists to accurately and consistently evaluate and diagnose DR. Ubotica are working with fundus camera manufacturers to introduce IDRD to the marketplace by 2021.

FED4SAE Support

A significant challenge in the deployment of AI based solutions is in the supervising of decisions of the CNNs which are at the core of these systems. Through the FED4SAE program we are working with fortiss to integrate their Neural Network Dependability Kit (NNDK) into IDRD. Among other features, it supports neural network (NN) verification to ensure the results do not lead to undesired behaviours, test-case generation and quality metrics computation to increase model robustness, and runtime monitoring to check if a decision of a deployed model is supported by the training data. In this way we are introducing a level of confidence to IDRD that heretofore was not possible, giving us a credible Verifiable AI element to our offering. In addition, working with our FED4SAE partners fortiss, Intel and Blumorpho has had a very positive impact on both our technical and business capabilities.

Impact

Deploying our IDRD solution in fundus cameras will significantly enhance the role of both the cameras and the ophthalmologists who use the cameras in the eye care paradigm, with further advances anticipated with the introduction of detectors for other pathologies, such as Glaucoma and Hypertension, all based on the same underlying hardware. With the integration of the NNDK the product offering has been significantly enhanced, with its supervisory role on the decisions being made by the CNNs delivering a significant benefit to the integrators of IDRD.

INCOMING: smart Interface for COnnected MachinING operation

Challenge

In Industry 4.0, the manufacturing world is embracing information and communication technology (ICT) more and more. Applications range from production management to smart material handling, from autonomous vehicles to sensor networks for machine monitoring. To gain the maximum benefit from adopting these systems, production machines must be directly interconnected with the factory ICT.

Making a network connection is only the start: making all machines, sensors and systems interact and play together like a well-tuned orchestra is the actual challenge. Efficient ways must be found to specify these interactions, since conventional programming would be too time consuming. Another significant aspect of this challenge is the need to bridge the technology gap and integrate legacy machines with no or only rudimentary digital networking capabilities.

Solution

INCOMING will develop a new integrated and interoperable solution to connect different types of machines through an open system that can be easily scaled and modified. This will be composed by an open source infrastructure for distributed industrial process measurement and control systems based on the 4DIAC platform, with a STM32 based solution as production machine add-on hardware.

To demonstrate the system functionality, this solution will be installed in three turning machines, with three different generations of machine control technologies. Machine operators and production managers will be able to monitor and control status and parameter data from these machines, enabling them to increase the overall equipment efficiency (OEE) by discovering problems and process inefficiencies faster.

FED4SAE Support

INCOMING will be carried out thanks to support by fortiss for the 4DIAC implementation and to support by ST Microelectronics for the chosen prototype STM32MP1 boards with X-NUCLEO expansion. Business innovation coaching is supported by Blumorpho.

Impact

INCOMING will generate impact in several ways. The first one is to establish a digital infrastructure that allows Zannini to gather operational status data from tool machines, which in turn can be used to improve production management processes, with the goal of increasing the Overall Equipment Efficiency (OEE) of these machines by an amount between 5% and 10%. The second reason is to permit Zannini to gain knowledge about the advantages of open software and hardware platforms for machine connectivity over commercial ones, so as to gain a competitive advantage in a fast changing market. Furthermore, the know-how that will be generated during INCOMING will also introduce the possibility for Zannini to create new services or even products for other companies.

Smart-Tunnel: Improving automatic incident detection in tunnels

Challenge

There are thousands of vehicular tunnels in use worldwide and that number is expected to continue to grow to alleviate the impact of congested traffic. These tunnels mainly deploy CCTV cameras which are wired back to a central control room where human operators monitor the flow of vehicles to ensure the safety of the tunnel and its users. A known problem is that human operators struggle to scrutinize large numbers of monitors over extended periods of time due to cognitive overload.​

ISSD has developed an Automatic Incident Detection (AID) system and deployed this to multiple sites around Turkey to aid the human operators in their monitoring tasks. These systems analyse parallel video streams from multiple cameras in real-time and provide automated alerts to the operators in the event that an issue is detected which they may need to act upon. This makes their job easier, more accurate and efficient, and makes the tunnel safer for its users.​

The current systems utilize traditional Computer Vision image processing algorithms which suffer from known limitations of this approach. Thus, the challenge is to improve the system accuracy, reliability and performance while extending its capabilities and ideally reducing the hardware costs.

Solution

Neural networks have shown excellent performance in computer vision related tasks. Thus, the new AID system design will utilise neural networks to achieve the desired goals. By augmenting the current AID solution, it is anticipated that the final system will be able to outperform the current system, detecting and tracking the movement of the various vehicle types or pedestrians with higher accuracy and better performance. To achieve this it is planned to use Vision Processing Units (VPU) that are dedicated to video image processing instead of the current reliance on CPUs. This will provide better performance, reduce the system hardware costs and contribute to a scalable architecture.​

FED4SAE Support

Through the EU funded FED4SAE Program, ISSD will be able to apply its many years of experience in tunnel monitoring systems, partnering with fortiss, BLUMORPHO and Intel to deliver this new product.​

Partners will provide ISSD with business innovation coaching and contribute their expertise in product development. Fortiss will contribute technically via their Neural Network Dependability toolkit to reduce the uncertainties inherent in the operation of artificial neural networks. The desired solution is targeting the Movidius Myriad X VPU hardware – the most advanced VPU from Intel which features a Neural Compute Engine with a dedicated hardware accelerator for deep neural network inference. ​​

Impact

The prototype system will be deployed and validated at one of ISSD’s existing road tunnel customers in Turkey. The solution will enable ISSD to have a stronger product offering to improve tunnel safety solutions deployed by authorities. The solution can be deployed to both new and – as a retrofit – to existing tunnel control centre infrastructure and thus bring the benefits of AI enhanced computer vision to these markets.

Video

Time4PS: Fully integrated development tools for partitioned systems

Challenge

The main goal of the project is to provide a complete set of tools that covers all the phases of the life cycle of the product: from the system design to the final configuration file of the partitioned system.
Goals

• Connect our customers modelling tool with Time4Sys.
• Give Time4Sys the ability to define partitioned systems.
• Connect Time4Sys and Xamber so the timing performance verification is automatically done in Xamber.
• Demonstrate de capabilities of the partitioned framework developed by means of an application experiment in the space domain.

Solution

The architecture proposal can be seen in figure below, where three tools are connected to define the complete system. A modelling design tool (EEI/XPM) that is used to model the system and generate the deployment and automatic code of the applications, Time4sys to define temporal information and a configuration tool to obtain the static configuration file for XtratuM (Xamber).

FED4SAE Support

Time4PS will contribute to scientific progress of software engineering for mixed criticality embedded real-time systems and it will highlight the increase in quality and effort savings obtained by using an integrated software development lifecycle supported by the appropriate tools. The close interaction of fentISS with FED4SAE partners, notably the Thales Group research team, will enhance the project impact in the scientific community.

Impact

The participation in Time4PS will enable fentISS customers to use more sophisticated tools for modelling, analysis and integration of their system and it will give fentISS the ability to develop further support of their products. Also, it will empower fentISS to gain global competitiveness in three ways. Firstly, it will allow fentISS to enter in customers who are already using Time4Sys. Secondly, it will improve the company products position in large and complex projects. Finally, the project will offer new opportunities to fentISS to enter in adjacent markets such as aviation, railways or automotive.

Runtime Architect – RATE

Challenge

Today, one of the major challenges in the Performance Engineering of real-time systems is the integration of design models and runtime aspects. The timing behavior at runtime has to be matched with the design in order to identify the timing failures in design and deviations from the real-time requirements.

Many tools exist for tracing the execution/communication and performing measurements of runtime properties. However, these tools do not allow the integration with system design models – the most suitable level for engineers for performance verification and optimization as well as decision-making.

Solution

RATE will allow the architect performing a continuous system performance engineering cycle between design and runtime, thus ensuring the quality of the running real-time system while reducing the design and development efforts and costs, and getting valuable feedback that can be used to boost the productivity and provide lessons-learnt for future generations of the product.
The following objectives will be fulfilled:

1. Verify if timing requirements are met in the system execution (CPU) or communication (network/bus) and identify potential timing bottlenecks.
2. Help the architect understanding the system timing behaviour based on numerical and graphical statistics for the system execution/communication.
3. Perform consistency check between runtime and models to validate timing related assumptions taken at the design phase.
4. Help the architect correcting timing errors and exploring design alternatives at the design phase based on the system runtime behaviour before investing time and efforts in implementing and testing.

FED4SAE Support

RATE will integrate the FED4SAE technology Time4Sys. By relying on the Time4Sys design model and the Time4Sys trace model, RATE will automatically benefit from current and future connections to Time4Sys of the various existing model-driven performance engineering tools such as design tools, scheduling analysis and simulation tools as well as tracing tools all at once. This will guarantee high flexibility and add a valuable agnostic character to RATE since it will be possible to easily integrate it in any runtime, design and any scheduling verification environment.

Impact

Runtime Architect gives to the architect access to knowledge on the system timing behavior based on the processing of runtime traces, thus allowing him to easily perform correction during validation and before delivery. Runtime Architect will allow reducing the design efforts (~15% gain estimated) while improving development efficiency as well as validation speed (~30% faster timing validation).

Through the integration of the Time4Sys technology, it will be easy to use Runtime Architect in any runtime, design and timing verification environment, thus offering to our customer a seamlessly integrated solution that will deliver a full, round-trip support for runtime traces analysis, design modelling and timing verification. Performance engineers will be able to rapidly integrate runtime aspects collected from traces into their design models. They will be able to start from runtime traces and then change the system configuration virtually to easily predict the performance impact of modified timing properties of tasks, schedulers or hardware. Runtime Architect will enable performance engineers to quickly iterate their designs as many times as they want, for both new developments and evolutionary extension and optimization of existing systems.