AI, MOBILITY AND INTELLIGENT IoT

dr. Luka Bradeško

Artificial intelligence

Laboratory

Artificial Intelligence Laboratory ~60 people working in various areas of artificial intelligence (m. learning, data m., semantic technologies,

computational linguistics, decision support, e-learning, mobility), tech transfer and dissemination

Sources of financing:

EU Projects

Industry projects

Academic partners:

Carnegie Mellon, Cornel, Stanford, MIT, Uni.

Maryland, KIT, UCL, W3C

Selection of FP6, FP7 and H2020 projects H2020 Optimum – Proactive influence on mobility users towards green transport

H2020 PrEstoCloud - Proactive Cloud Resources Management at the Edge for Efficient Real-Time Big Data Processing

H2020 euBusinessGraph - Enabling the European Business Graph for Innovative Data Products and Services

H2020 BigDataFinance - Training for Big Data in Financial Research and Risk Management

FP7 IP ACTIVE – Enabling the Knowledge Powered Enterprise

FP7 IP COIN – COllaboration and INteroperability for networked enterprises

FP7 IP EURIDICE – Inter-Disciplinary Research on Intelligent Cargo for Efficient, Safe and Environment-friendly Logistics

FP7 NoE PASCAL2 – Pattern Analysis, Statistical Modeling and Computational Learning

FP7 NoE T4ME – Machine Translation & Multilingual Information Retrieval

FP6 IP ECOLEAD – European Collaborative Networked Organizations Leadership Initiative

FP6 IP SEKT – Semantically-Enabled Knowledge Technologies

Artificial Intelligence Laboratory HELPING THE INDUSTRY WITH STATE OF THE ART RESEARCH AND DEVELOPMENT:

Accenture Labs: enterprise search, process modelling, visualization Bloomberg: anomaly detection, failure prediction, bond trading and calculation, web analytics, activity

prediction, x-recommendation, AIQ British Telecom: alarm root cause analysis, failure prediction, anomaly detection, traffic prediction,

freight transport management and optimization Google Labs: opinion detection in Google News IBM Watson: media understanding and news formalisation Microsoft Research: product expert detection, complex data visualization, text mining consultancy New York Times: real time web analytics, trend detection, visitors modelling Siemens: “Factory 4.0” consultancy UNESCO: Open Educational Resources processing, learning path calculation, micro-credentials UN FAO: information integration Wikipedia: opinion detection, content analytics Post of Slovenia: semantic and analytics services for e-govenment, cognitive logistics Kolektor: Factories of the Future analytics, prediction, planning and optimization environment AdriaMobil: Intelligent Camper

Artificial Intelligence Laboratory TRANSFER THE RESEARCH INTO PRODUCTS

Spin off companies and NGOs:

Solvesall d.o.o.: Smart Mobility and IoT. Smart motorhome.

Quintelligence: Qminer, predictive analytics, complex space modelling and failure detection

Qlector: Operational analytics, production planning, …(Joint venture with Kolektor)

Eventregistry: Global pulse through media analytics

JEMS: IoT supported waste2fuel technologies

Knowledge4All foundation (London)

AI4Good (New York)

Big Data for Social Good (Chicago)

Opening Up Slovenia (open education)

Solvesall MOBILITY SERVICES. AI, MACHINE LEARNING, SOFTWARE AND HARDWARE DEVELOPMENT

○ 2 PhDs

○ 2 masters

○ 1 designer

Team of 5

➢ Used to work together for 1 year at

Solvesall and 5+ years at AI-Lab,

Jožef Stefan Institute.

➢ Previous experience includes

working in research and industry,

incl. Bloomberg L.P., and startup

developing NL/AI based Personal

Assistant.

Currently working with:

➢ Adria Mobil (motorhomes, caravans)

➢ Adria Dom (mobile homes/houses)

➢ LPP (Ljubljana bus operator)

What is E-Mobility?

The same as mobility + electric transportation options

Mobility != Transportation

Mobility = Having affordable and safe transportation options that one

can use and count on. Including walking.

Cars, bicycles, e-scooters, public transport, …, …

What E-Mobility consists of?

➢ The same as mobility and a bit extra (technical): ➢ Users

➢ Vehicles

o Hardware

o Software

➢ Infrastructure

o Roads, pavements

o Signalization

o Software + Services

o …

➢ Charging infrastructure

o Charging stations

o Grid

What E-Mobility consists of?

➢ The same as mobility and a bit extra (technical): ➢ Users

➢ Vehicles

o Hardware

o Software

➢ Infrastructure

o Roads, pavements

o Signalization

o Software + Services

o …

➢ Charging infrastructure

o Charging stations

o Grid

AI in Mobility

➢ Autonomous vehicles

AI in Mobility

➢ Autonomous vehicles/driving

➢ Predicting most probable destinations when turning on navigation

➢ Predicting electricity consumption of the grid (charging needs)

➢ Predicting parking space availability

➢ Predicting docking station availability for e-scooters

➢ Traffic signs recognition

➢ Accident detection from traffic cameras

➢ Start/Stop system optimization (learn to prevent stop/start within 1s)

➢ Traffic measurement and prediction (floating car data + loop sensors)

➢ Map-Matching from GPS data

➢ Learning User Mobility patterns

➢ SOC detection…

AI in Mobility

➢ Autonomous vehicles/driving

➢ Predicting most probable destinations when turning on navigation

➢ Predicting parking space availability

➢ Predicting docking station availability for e-scooters

➢ Traffic signs recognition

➢ Accident detection from traffic cameras

➢ Start/Stop system optimization (learn to prevent stop/start within 1s)

➢ Traffic measurement and prediction (floating car data + loop sensors)

➢ Predicting electricity consumption

➢ Map-Matching from GPS data

➢ Learning User Mobility patterns

➢ …

DATA

Mobility Software Infrastructure

➢ Data Realtime (Slovenia is a good test-bed) ➢ 430 road loop sensors

➢ Traffic Events

➢ Traffic Tweets

➢ JP LPT parking data

➢ Wind Sensors (Primorska region)

➢ Weather

➢ Floating car data (270 LPP busses, user mobile applications, ~40 rental motorhomes, 21

taxi’s)

➢ Bicycle sharing station data (BicikeLJ)

➢ Traffic cameras

Mobility Software Infrastructure

➢ Data Realtime (Slovenia is a good test-bed) ➢ 430 road loop sensors

➢ Traffic Events

➢ Traffic Tweets

➢ JP LPT parking data

➢ Wind Sensors (Primorska region)

➢ Weather

➢ Floating car data (270 LPP busses, user mobile applications, ~40 rental motorhomes, 21

taxi’s)

➢ Bicycle sharing station data (BicikeLJ)

➢ Traffic cameras

Mobility Software Infrastructure

➢ Software Services ➢ Mobile libraries for tracking

➢ Data Connectors and Watchdogs

➢ Analytics services

➢ Traffic Prediction

➢ Bicycle station availability prediction

➢ Personal Mobility Patterns

➢ Car-sharing Match-making

➢ Routing and Map Matching

➢ POI relevance and star rating assessment

➢ Trip detection

➢ Location visit detection

➢ Public transport vs personal transport classification

➢ Predicting next user location

➢ Group Analytics ➢ Frequent routes of the whole group

➢ Frequent locations of the group

➢ Modalities

Traffic Flow prediction

Bicycle (charging) station availability prediction

Detecting traffic flow for the shared bikes in LJ (not electric)

Demo application for data collection using the library

Map-Matching and route detection

Visit detection and travel mode classification

Automatic Home and Work detection

Automatic Home and Work detection

Predicting the next location

Frequent Personal Locations

Frequent Routes

Route Modality Statistics

Real-time group/fleet view

Fleet/Group Frequent Locations

Fleet/Group Frequent Routes

Examples of usage we tried

➢ Proactive push of traffic participants towards green transport

➢ Cost of car ownership vs using public infrastructure, taxis and

car sharing

➢ Increase car-pooling pool with the future travels

➢ Learning of points of interest for specific groups/fleets

➢ Calculating personal safety index

Connected (Smart) vehicles

➢ Connected Smart Motorhome Prototype

MACH (MOTORHOME AI COMMUNICATION HARDWARE)

➢ RV focused platform for

○Remote control, predictive analytics and mobility services

➢ Connects and control all appliances in the RV

➢ Integrates 3rd party services

○Camping, navigation

➢ AI, ML:

○POI influenced by sensor states

○Automatic POI learning

○Resource consumption predictions

○ Experimental: autonomous behavior

THANK YOU!

Questions?