Descripción del proyecto
Nuevas herramientas para ayudar a los expertos en logística a afrontar el reto del último kilómetro
En el mundo de la logística, el problema de la entrega de último kilómetro es uno de los retos más importantes, ya que crea atascos que aumentan los costes y la ineficacia. Los métodos tradicionales de optimización de rutas se tambalean ante las perturbaciones en tiempo real, lo cual deja a los expertos en logística ante un viejo dilema. El equipo del proyecto SmartDelivery, que cuenta con el apoyo de las acciones Marie Skłodowska-Curie, aprovechará la sinergia de aprendizaje automático y el internet de las cosas para abordar la entrega de último kilómetro. En concreto, introduce una novedosa arquitectura de «hardware» y «software» que utiliza datos de vehículos en tiempo real para mejorar continuamente los algoritmos de enrutamiento. Además, un innovador método basado en internet de las cosas asigna dinámicamente rutas a los conductores, guiados por un parámetro único del «sexto sentido». Un módulo de aprendizaje automático predice el algoritmo heurístico y metaheurístico óptimo para perfeccionar la ruta.
Objetivo
Scientific advances in recent years have brought to light a series of potentially disruptive technologies in the ICT landscape. They are becoming, and will increasingly become, key enabling technologies for the development of applications and services designed to improve the quality of life of citizens and make processes more efficient. Among these, we can identify some which research has recently focused on with particular attention: Machine Learning and Internet of Things. In this project we propose a combined use of these two technological enablers to solve one of the main issues which all logistics experts have to face: the problem of optimising the last mile delivery (LMD). LMD is a crucial step of the entire delivery process, as it causes bottlenecks and is typically the most costly, problematic and inefficient part. Improving the LMD process in terms of route optimisation using classic approaches is difficult: static algorithms are not suitable, and even heuristic algorithms do not find high-quality solutions, as they do not consider several factors such as unpredictable real-time events which may occur. To address these challenges, a novel hardware/software architecture which exploits real-time vehicles’ positions to continuously improve performances of the routing algorithms is proposed, together with a new IoT-based methodology to automatically/dynamically assign routes to drivers based on the values of a defined “sixth sense”parameter. A ML module will predict the best among a chosen portfolio of different heuristics/metaheuristics algorithms to optimise the route.
Ámbito científico
- natural sciencescomputer and information sciencesinternetinternet of things
- natural sciencescomputer and information sciencessoftware
- natural sciencescomputer and information sciencesartificial intelligencemachine learning
- natural sciencescomputer and information sciencesartificial intelligenceheuristic programming
Palabras clave
Programa(s)
- HORIZON.1.2 - Marie Skłodowska-Curie Actions (MSCA) Main Programme
Régimen de financiación
HORIZON-TMA-MSCA-PF-EF - HORIZON TMA MSCA Postdoctoral Fellowships - European FellowshipsCoordinador
37008 Salamanca
España