Hjalmar Delaude, Jamente Cooper, Sivakumar Pillai, Istvan Barabasi

Slides:



Advertisements
Similar presentations
Research Directions for the Internet of Things Supervised by: Dr. Nouh Sabry Presented by: Ahmed Mohamed Sayed.
Advertisements

Ch.2. Ubiquitous IoT Applications
© Siemens AG, 2002 s CP RS Agenda The Role of IT for Accident-free Driving Interaction with driver’s physical condition Interaction with the roadside environment.
Smart Smart Contracts AI on the blockchain 1 | Neural Networks on Ethereum’s Distributed Virtual Machine.
Internet of Things (Ref: Slideshare)
AppulateConnect April, 2016 The Power of Innovation
The Network Aware IoT Service at Edge Guoxi Wang.
Verizon Intelligent Track and Trace: Serialization and Cold Chain
Personal Home Healthcare System for the Cardiac Patient of Smart City Using Fuzzy Logic Shijia Liu.
Principles Identified - UK DfT -
Healthx Analytics: Turning Data Into Actionable Insights
Safety, Security & Savings
Network of Vehicles By, Adarsh R S.
Intelligent Transportation System
@Japan Smart City Trip February 24-28, 2017
About Me Name: Yaokai Feng, from Kyushu University
Blockchain Infrastructure for e-Science
Emerging Applications and Platforms in Embedded and Realtime Systems
5/13/2018 1:53 PM © Microsoft Corporation. All rights reserved. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN.
Connected Maintenance Solution
Intelligent IVI with AI
Sponsored by J&J (PSV) Magyarország Kft.
Deep Shift: 21 Ways Software Will Transform Global Society
IoT at the Edge Technical guidance deck.
Reducing the Risk of Injury
Researcher and Lecturer Assistant at FCI Assiut University
1st Draft for Defining IoT (1)
Connected Maintenance Solution
Reducing the Risk of Injury
doc.: IEEE <doc#>
Azure-Powered Contract Management Solution Turns Business Contracts into Smart Contracts “The scalability of the Microsoft Azure cloud platform’s infrastructure.
How Smart Networks are Changing Corporate Networks
Blockchain beyond cryptocurrencies
(1888 PressRelease) First Decentralized Private Communication Solution
Blockchain Adrian Zaragoza.
{ BLOCKCHAIN Technology. BSEtecBSEtec is a digital solution provider company which offers the best service with the implement of the latest technologies.
IoT at the Edge Technical guidance deck.
Public sector services
Ottawa AV Innovation ‘201’ Changing the way traffic moves
From “Cash on the internet” to “Digital Gold”
Artificial Intelligence Changes the Security Landscape
Securing the Internet of Things: Key Insights and Best Practices Across the Industry Theresa Bui Revon IoT Cloud Strategy.
Carl Data Solutions Collects Utility Sensor and Meter Data to Provide Advanced Reporting, Alarming, and Analytics with Microsoft Azure MICROSOFT AZURE.
Artificial Intelligence in Manufacturing
Distributed Ledger Technology (DLT) and Blockchain
Blockchain Alexander Prenta 9/27/2018.
Genes to Function to Therapeutics
Internet of Things.
Smart Learning concepts to enhance SMART Universities in Africa
Big Data and IoT FTG-07.
DIGITAL DIVIDEND : TECH FOR GOOD
DATS International Portfolio.
Internet of Things Stay Relevant in Digital Era
The Sense Add-On Module for MarineCFO’s Vessel 365 Solution Uses Microsoft Azure Services and IoT Technologies to Provide Advanced Data Analytics MICROSOFT.
IoT, Tekoäly ja wearables
Mathias Johanson, Jonas Jalminger Boel Nelson, Tomas Olovsson
Driving sustainable food production with Precision Agriculture.
D A L I Deep Artificial Learning Intelligence
Developing Vehicular Data Cloud Services in the IoT Environment
Global megatrends (relevant for our business)
Sachiko A. Kuwabara, PhD, MA
Blockchain Technology
Welcome to The World of Internet of Things
CELTIC-NEXT Event 20th June 2019, Valencia
© 2016 Global Market Insights, Inc. USA. All Rights Reserved Fuel Cell Market size worth $25.5bn by 2024 Low Power Wide Area Network.
Utilizing the Network Edge
The Intelligent Enterprise and SAP Business One
Explore Txs, block, blockchain in Bitcoin
IoT and Supply Chain Risk Management
Presentation transcript:

Hjalmar Delaude, Jamente Cooper, Sivakumar Pillai, Istvan Barabasi How Blockchain and AI accelerate the application of IoT technology to keep drivers’ safe on the road Hjalmar Delaude, Jamente Cooper, Sivakumar Pillai, Istvan Barabasi

AI, ML, IoT and Blockchain offer new insights to car and driver interactions Advancements in artificial intelligence (AI) and machine learning (ML) help implement smart technology into cars and provide new ways of interactions with the drivers. Internet of Things (IoT) technology implemented within cars enables collecting rich data about the road location, traffic, driver’s behavior and overall car and driver interactions. Trusted communication with the driver and IoT device(s) within the car can be established using Blockchain technology. Blockchain enables trusted delivery of detected anomalies and/or alerts to critical service providers. This technology is readily available to students and faculty, making easy implementing prototypes and performing research in and for problems areas of their choice

Background According to Centers for Disease Control and Prevention (CDC): in 2015 over 10,000 people died in alcohol impaired driving accidents, which accounts to about 30% of all fatal traffic accidents within the US. In the same year, about 1.1 million drivers were arrested for driving under the influence. CDC’s Behavioral Risk Factor Surveillance System (BRFSS) provides data and reports regarding health-related risk behavior for 50 states and District of Columbia. Today this data is collected mainly based on surveys. Using IoT technology, BRFSS could collect real-time data about drivers and analyze risk factors, trends. IoT data would enable BRFSS to real-time assess risk levels using AI and subsequently help implement solutions that can mitigate and lower them. Important factor for such AI solutions is establishing trust between source of data and back-end cloud services and protecting drivers’ privacy. Key requirement for AI solutions is how to establish personalized service with the drivers, but also anonymize the same data when providing large scale reports and statistics about the drivers. Detecting critical conditions affecting drivers ability to safely operate a vehicle can help save lives.

Raspberry PIs as private Ethereum nodes We have chosen Ethereum as our choice to implement private blockchain, smart contract and trusted transactions Blockchain technology was introduced by Satoshi Nakamoto (2008) in his paper “Bitcoin: A peer-to-Peer Electronic Cash System” (https://bitcoin.org/bitcoin.pdf). The problem he solved in his paper was the constitution of trust in a distributed network. Ethereum is an open-source decentralized platform based on blockchain, implementing smart contract (scripting) functionality. According to Wikipedia, Ethereum implements a modified version of Nakamoto consensus via transaction based state transitions. We have downloaded Ethereum (geth) from https://geth.ethereum.org/downloads/ We followed best practices and technical guides from chainskills.com

Solution outline for driver condition detection Cloud Analytics Service Alert as sent as trusted transaction via blockchain Heart Rate monitoring using Fitbit ChargeHR Private Blockchain Sleep condition sensing via Fitbit ChargeHR Central Computer Critical condition detection: Driver falling asleep Driver heart condition Reckless driving GPS location and accelerometer sensor data capture via BerryGPS-IMU We have collected sensor data using Fitbit and BerryGPS-IMU devices Sensor data was sent to a Raspberry Pi (RPI), where conditions are detected using programs and algorithms. The RPI can communicate with cloud services for analytics, such as IBM Driver Behavior The RPI initiates trusted blockchain transaction to a central computer Alert is received and processed (forwarded to emergency services)

Implementing AI and ML in context of IoT devices When using AI and ML with small IoT devices, implementing solutions requiring intensive processing becomes an engineering problem Analytics can be performed in the cloud and results sent to IoT device Other approach is to implement lightweight engines on a device and perform analytic at small scale A recent third option is to pre-train models, reduce them and apply download to small IoT devices Recent IoT at edge technology available to us is the Intel Movidius Compute Stick This stick enables implementing machine learning at the RPI The model (such as for KNN classifier) must be trained on a larger computer compacted and downloaded to the stick

Any questions?