Mr. Thomas Geraghty

Summary

Solving problems in computational fluid dynamics (CFD) is numerically intensive but amenable to 'divide and conquer' parallel algorithms. In this project, the most numerically expensive elements of the solution algorithms which typically solve a large set of linear equations will be implemented at a hardware rather than a software level. This will reduce the run times by a factor of between 100 and 1000, when compared to current techniques. To achieve this an FPGA based numerical accelerator comprising multiple parallel computational units (CU) will be implemented to accelerate the solution of these problems. The accelerator will be a co-processor to a workstation and will operate transparently to the user. It will feature 'on the fly' hardware re-configurability wherein the hardware may be customised to meet the specific demands of a problem.

Funding Agency

Enterprise Ireland

Programme

Research Innovation Fund - €145,000

Person Months

24

Summary

Delay tolerant networking covers constrained networks where unavoidable communication delays (hours/weeks) are such that normal "chatty" Internet protocols (especially TCP) cannot be used. The "canonical" examples of delay tolerant networking are interplanetary and sensor networks. This project builds upon ongoing work in interplanetary networking by applying this work to build, from commercial off-the-shelf (COTS) products and freely available technology, an innovative architecture which allows novel sensor network applications to be easily and cheaply deployed. The project emphasises the evidentiary (security) requirements that must be met for such data to be credible. The result is a proof-of-concept, secure, delay-tolerant, sensor network.

Funding Agency

Enterprise Ireland

Programme

Research Innovation Fund - €142,000

Person Months

24

Summary

The objective of the project is to develop and field test a low-cost system for static and dynamic axle load measurement for real-time weigh-in-motion for heavy goods vehicles (HGVs). The system will be integrated with a commercial 'off-the-shelf' (COTS) unit that provides GPS capability and is interfaced to a Geographical Information System (GIS). The integrated system will provide real-time data on loads transmitted to road surfaces.

Funding Agency

Enterprise Ireland

Programme

Research Innovaton Fund - €146,000

Summary

Currently the performance of individual science/engineering workstations is limited by silicon I.C. technology and CPU architectures. This means that to solve complex numerical problems in aeronautics, civil and chemical engineering product designers must use multiple workstations or a high-performance computer. Either option is expensive in terms of both capital and maintenance, restricting the amount of computer modeling which can be performed by an individual scientist or engineer. This proposal aims to deliver 30-60 double-precision GFLOPs (Giga Floating-Point Operations per second) of computing power to the desktop using FPGA technology rather than custom silicon (to reduce design and deployment costs) for the price of a high-end graphics card, in an era when a single Pentium 4 processor can deliver 6 GFLOPs at 3GHz, and realistically could deliver twice this in the next generation. To make this computing power available to engineers and scientists the proposed hardware accelerator will be complemented by software libraries for C, C++ and Fortran as well as for commercial packages (e.g. FLUENT, ANSYS) removing the need to rewrite legacy software.

Funding Agency

Enterprise Ireland

Programme

Commercialisation Fund - Technology Devleopment - €350,000

Summary

Dev-SeNDT is geared towards the "meta-objective" of developing SeNDT to the point where it will be licensed and we aim to negotiate such licenses. The specific, verifiable project objectives to enable this are: 1. Running significant pilots: water quality monitoring on Trinity and Carr's lakes in Co. Cavan, and noise monitoring in an urban area to demonstrate that the technology meets the challenges posed more effectively than competing approaches. a. One highly specific and testable subsidiary objective here is to run a lake water quality monitoring network for an entire winter with only solar power and without exhausting batteries. b. Demonstrating that the SeNDT software and technology platform is applicable in diverse environmental monitoring applications by showing that both boats and busses (or equivalent) can be used as data-mules without unnecessary software or hardware changes. 2. Demonstrating that SeNDT is ready-to-license for existing or new product lines. a. Integration of the SeNDT technology with a commercial-grade GIS and database back-end - necessary for practical application development and deployment. b. Decreasing the unit cost of SeNDT nodes via a cost-reduction re-design of the hardware to take advantage of current products, and refining the system requirements to better fit with our more mature understanding of the real-world application needs. 3. Researching whether a simple webcam based approach can be used to ameliorate the maintenance and robustness problems with chemical sensors in lake water quality applications. 4. Promoting the use of, and continuing to add to the state of the art in, delay tolerant networking.

Funding Agency

Enterprise Ireland

Programme

Commercialisation Fund - €343,000