Get the best ... project!?

Several project variations are possible, but, of course, you can develop a variation on this project to help you choose the best project for yourself ! ;-) Although the range of options are broad and may appear somewhat unfocussed, the aim is to inspire and stimulate you to find a project in keeping with your interests and/or background. However, this may then incur the attendant drawback of your then having to formulate more realistic concrete project goals, in view of data availability.


Get the best ... system!?

The aim of the project is to monitor and log performance and response of any system to a stimulus, with the aim of characterizing and optimizing a desired response. Target systems may range from computing, through ecology, agriculture, sports to health monitoring and management of industry or society, and apply to individuals or groups, such as herds, or groups with closer mutual dependence and interaction, such as teams, or even societies, if you are socially (or cynically) oriented. Optimizing a response may involve stabilizing a system, as in health applications or maximizing some metric as in sports or production.
Suitable implementations or applications in finance, or sport (or both e.g. fantasy leagues, horses, dogs) may spark commercial interest and/or support; (or financial ruin and debt given the current global economic outlook.!)

Approaches may be data or theory driven, and range from Artificial Intelligence approaches through classic control theory, or any combination.

Sample project applications : Cluster / network management & load balancing. Economic / trade links & (im?)balance of payments! (see another project on Cash Crashes..) Animal / herd monitoring .. e.g. milk yield vs. feed, housing; Crop yield vs. management schedule (prepare, sow, treat, reap), weather, fertilizer; racehorse performance vs. race schedule, jockey, race strategy, feed and training regime; team performance vs. match schedule, player mix, opposition; political performance vs. extreme media exposure, voter sentiment, historical backlashes; Irish economic performance vs. rugby team wins, political maturity and integrity (immeasurable!?) and EU funds!; public health vs. health policies: e.g. smoking ban screening, immunization programmes; climate vs. sunspot activity, peak oil, political power balance etc.

Note that it may be difficult to obtain data for some of these, but the range of suggestions may encourage you to think of another application of interest to you.



Flash mobbling – or mo(re)-bling!?

Develop a smartphone app, in which phones act like a flash mob, spontaneously bursting out in synchronized music or song! Clearly the phones need to synchronise, with the same music, or with different scores, parts, voices or instruments, and could be programmed to produce curious immersive aural experiences, including anti-phase, which gives a peculiar ethereal effect.

Implementation would normally require, either an ad-hoc or WiFi network, for high-bandwidth transfer of music and fast synchronization, although this could also be attempted over a normal cellnet, but may not function quite so well.

Variations and other applications are at the discretionary creative imagination of the student, but may include centralized (orchestral conductor) or fully distributed control (seisiún / session), with easy admission or exit from the group.

This is a project for someone with a passion for mobile networked music making! A knowledge of mobile networks and programming would clearly be an advantage.



Life_Guard.

Several smartphone apps exist (mostly for Android : Daily Voyager, Autoguard...TomTom for iPhone) which use standard facilities such as GPS , accelerometer and camera to record and retain interesting parts of a journey, but no significant postprocessing is currently facilitated.

The project should support uploading of recordings to a website, permit retrieval of time and location of ‘interesting’ events, and attempt to spot any abnormal frequency or sequences of occurrences by time or location. This could help alert both the public and those involved in taffic management to the existence of hazardous places, times or behaviour.

Recordings can be classified by the user during upload, to ease project implementation and detection of ‘interesting’ events. Such classifications can be ‘normalised’ over time to relate the individual sensitivity of the specific user to that of the general public perception.

The project offers scope for expansion, or several related projects. For those with an image analysis or HPC passion, number plate recognition or scene analysis may help automate the process to some extent.


Tell-tale signs ... science!?

Social media monitoring viral spread of political sentiment, economic recession, or health epidemics!.. opinion mining, sentiment analysis various projects possible Combine social media and modeling to predict contagion affecting the outcome of the forthcoming presidential election, or the spread of disease, or whether banks and currencies will fail based on international exposure involving trade, investment and other political factors.


Cash crash!?

A specific variation of the “Tell-tale signs … science!?” project, where economic shock and contagion is modeled simply (and the simpler, the better able to see the source) by sentiment (and consequently trade) propagation throughout the global economic system modeled as a network (or networks for communication in trade and culture etc.).


Public health monitoring through social networking.

A specific variation of project no. 40 ; other variations still possible.


The aim is to monitor the spread of illness by analysis of various social networking sites and compare this with statistics from other sources.



Plot an educational course!?

(this is only by way of example, Dr. Colin McCormack supervises projects in this area)

The basic aim is to develop a web-enabled software system which would be of use to students selecting modules in keeping with their aspirations and requirements, and also of use to staff designing courses, or advising students. The application should logically validate and generate a graphical representation of any course, which is merely a collection of modules, subject to constraints obtained from the UCC online calendar. Typical constraints include: required mixes of essential core and optional elective modules for various programme requirements; total credit count for the year and balance per semester; pre & co-requisites within and across all semesters and years, and timetabling clashes. It is hoped this will reduce errors in specifying and interpreting calendar course requirements, which are frequently just given as module codes, without accompanying text. The graphical output should distinguish between essential core and optional elective modules across various streams, displaying all options clearly to differentiate and ease selection. All options (year, stream, module) in the graphical output should be hyperlinked to permit browsing for more details.

It would be an added advantage, if it were to generate a lecture timetable specific to the module mix from the central university timetable, and were also to produce a navigation guide from one class venue to the next to ensure smooth starts at the beginning of each term.

This project should be suitable for a student with competent broad programming skills, with basic web development skills, text parsing and graphical display capabilities, and a good appreciation of clear GUI principles.


Plot your course … for travelling…!?

Clearly a variation on the previous project exists in the form of planning a travel itinerary using published online scheduled travel timetables. A proof of concept sufficient for a project may include Mizen to Malin (or as near as possible) by public transport, with typical constraints such as least cost, time or changes. Of course, this could be extended to the global transport network. Obviously, similar things have been done before, but you could add your own unique creative touch!