moveON


Metaheuristics, Holistic Intelligence, and Smart Mobility

TIN2014-57341-R (2015-2018)

TIN2014-57341-R

2015-2018

This project makes an ambitious proposal for focused research in challenges related to intelligent transport and smart mobility. We do it from the perspective of building new applications based in solver metaheuristic engines enhanced with methodologies and theories contributed by our team so as to exhibit "holistic intelligence".
We have recognized several gaps in present research related to intelligent transport systems (ITS) and also in the future of metaheuristics. In this project we make a structured approach so as to both make strong research and construct, at the same time, proof of concepts with interest for companies. Our ingredients for a successful proposal are:

  • A. A main goal of improving how we could measure, model and solve problems in ITS and smart cities. We address these three targets in all the project applications.
  • B. A clear set of requirements that will restrict all our algorithms and applications. We call it the "six pack", and they basically amount to analyze any algorithm's behavior from six points of view (and metrics) at a time: flexibility, self-adaptability, robustness, high dimensionality (scalability), efficiency, and compliance to city standards. Any algorithms and software exceling in all these are hard to built but worth in research.
  • C. We target 10 concrete research questions, six of them related to each requirement of the six pack, and four additional ones in the core of open challenges in optimization and learning: running algorithms on smart phones, getting new practical-theory in metaheuristics, combining parallelism+multiobjective+dynamic problem solving, and reusing knowledge and the built solvers in other fields (energy and intelligen trains).
  • D. We include 7 detailed goals, each one representing an open ITS domain of investigation: private transport (drivers, passengers in cars), public transport (bus and on-demand services), surface parking management, transport of residuals (routes acounting for slopes, forecasting of filling levels, smell sensors), optimization of traffic lights cycles, maps and alerts of the city, and cross-fertilization to other domains.
  • E. We will build a urban lab (uLab) and a computational lab (cLab) to have a lasting infrastructure in ITS of hardware and software tools for future research. The uLab accounts for all Hw&SW we will use to take data, process it, and test our services. The cLab relates to all SW&HW that we will use in the university premises oriented to ICT: a mixture of real data/devices and simulations (of sensors, actors, scenarios, and services).
  • F. A transference plan exists, involving several companies directly related to the applications.

With the mentioned components we want to explore in a holistic (integral) way both, academic advances and industrial interests. We consider in the project web services and apps, as well as we take care of the main players in ITS: citizens, companies, and city managers (the Ayto. of Málaga directly supports our studies).
At the same time we identify the scientific *starting points*: elementary landscapes, math oracles, systolic genetic search, swarm intelligence, a new concept on epigenetic algorithms, parallelism in manycores and GPUs, dynamic asynchronous distributed solvers, new time series models, manyobjectivization, proprioception...
In summary, we will offer data, benchmarks, applications, theory, software, and hopefully a fruitful transference so as to have here a serious proposal.

Goals


Advances in private car transport

Advances in public transport

P

Surface urban parking

Residual gathering of tomorrow

ImgSem

Advanced central control of red lights

ImgCar3

Know your city and Cross-fertilization to other domains

Publications

Problems

Software

NEO Software Area

Related Projects


The MaxCT Project

The RoadMe Project

The Carlink Project

The M* Project

Contact Us

Contact Info

Address: Bulevar Louis Pasteur, 35, 29071 Málaga (Spain)

Web: neo.lcc.uma.es

e-mail: eat (at) lcc.uma.es