Upload
others
View
4
Download
0
Embed Size (px)
Citation preview
PUBLIC
Thomas Kunert,
Senior Director Public Services, EMEA N, SAP
Digitization - what does it mean for public sector organizations? Season 3: Smart City and Smart Country Cases
2PUBLIC© 2019 SAP SE or an SAP affiliate company. All rights reserved. ǀ
Heidelberg
Trondheim
Austria
Berlin
Travelling Europe – Our Route Today
Internet of Things and Big Data Projects –all over Europe
Projects in unbelievably many shapes
Today‘s stations:
Optimization of Waste Management in Heidelberg
A bridge in Norway – and how tomaintain it better
A trip to Austria to see how they sourceand analyze traffic data for betterachievements…
How you can detect crisis situationsearly – to optimize resource allocation
3PUBLIC© 2019 SAP SE or an SAP affiliate company. All rights reserved. ǀ
City of Heidelberg – What was the issue ?
Overflowing Containers lead to unnecessary collections - waste of time and tax money Smell, noise, traffic jams Low service quality Quality of life
Did you know?
Lorries, that are commonly used in the waste
industry, consume around 70 litres of diesel
per 100 km.
… and aren’t overflowing recycle bins the opposite of a smart city?
4PUBLIC© 2019 SAP SE or an SAP affiliate company. All rights reserved. ǀ
The Digital Agency of Heidelberg and SAP kicked off an initiative with several partners
1st project: IoT enabled public glass recycling containers
partnering with disposal companies –benefits:
exact and predictable route planning higher efficiency and effectiveness the impact on cost
What to do with DGTL now?
Special sensor wireless network forData Transfer
Experts for sensors
Data provided is processed real time
‚Digital‘ Ingredients – What we put in….
Low Power
Wide Area Network
5PUBLIC© 2019 SAP SE or an SAP affiliate company. All rights reserved. ǀ
Pilot
Start: 10 containers in the city and the outskirtsof Heidelberg
equipped with highlyenergy efficient androbust sensors and radio
Receiver
IoT technology on theSAP Cloud Platform
Process Flow & Architecture
Data delivered
Filling – percentageof capacity
GPS data forlocation
Temperature
State of sensor egcovered or not
Delivery
Long Range WAN
Energy efficiency, cost effective
low data volume at low frequency
Dashboard
Map based
rules say whencontainer is full
Trigger process in backend
More analytics:
analyze trends and peaktimes – detect patterns
Insight into Action
Integrated intodisposal company‘sSAP waste andrecycling solution
Routes can bedynamically planned– demand driven
Potential: Fires can be detected early –via threshold values alerts canbe triggered and inform certaincontacts proactively
6PUBLIC© 2019 SAP SE or an SAP affiliate company. All rights reserved. ǀ
What it looks like….
7PUBLIC© 2019 SAP SE or an SAP affiliate company. All rights reserved. ǀ
Improve citizens Quality of Life – no more overfilled recycling containers and a cleaner city, better QoS
Was it worth the effort? Top Benefits.
More efficient city services –demand based collection on current sensor data
Less traffic impact by collection vehicles Less noise, less exhaust, less smell ( CO2! ) Less debris
City control the truck fleet better, unnecessary collections are avoided
City improves forecasting negotiate better contracts for the
services – better for the city and its citizens ( tax money )
8PUBLIC© 2019 SAP SE or an SAP affiliate company. All rights reserved. ǀ
…or just listen to the Mayor of Heidelberg
Heidelberg
Trondheim
Austria
Berlin
Travelling Europe: A Bridge in NO
You Need to Know
repair and replacement of infrastructure consumes tremendous amounts of time and money
Maintenance with traditional monitoring based on brief periodic inspections
public sector is looking for affordable and effective monitoring methods to ensure bridge safety
Can Digital technologies help?
10PUBLIC© 2019 SAP SE or an SAP affiliate company. All rights reserved. ǀ
The Case for Bridge Monitoring - Extending bridge life
“The technology ( … ) can maximize the bridge lifetime and start renovations at the optimal time,” … “If the
method can help postpone the replacement of a bridge by ten years, it will save the authorities large amounts
of money.”
Arild Christensen, oversees bridges and ferry ports for the Norwegian Public Roads Administration in central Norway
Aim: High reliability by
sensor-based real-time monitoring insight into the bridge’s behavior warning of structural failures
Situation Stavå bridge from WWII, designed for 20-ton freight loads, today: 60-ton loads Large vehicles only cross at reduced speed, one direction at a time.
How to maintain it better and keep it safe? The Digital Twin PoC of NPRA and SAP
11PUBLIC© 2019 SAP SE or an SAP affiliate company. All rights reserved. ǀ
Bridge monitoring – How? Analyses and Prediction
real world digital twin
external load
external load
virtual sensor
sensor
physical sensor
actuator
actuator
Sensor output provides data about the structure:
Symmetric behaviour Drifting of long-term
observations ‘strange’ behaviour sensors attached to
the bridge transmitted to base
station and twin– egvibrations.
input used to tune as-built model of the actual bridge structure
Threshold alerts - for drift and structural behaviour.
Development of the deterioration can then be estimated in the model.
Planned modifications eg repairs can be simulated using the tuned model to estimate the effect on the global structural behaviour
“We place virtual sensors and implement the failure that caused the collapse of the bridge,” ... “If you can find the correlations then you can predict failures before they happen.” Martin Hasle Product Delivery Manager, PEI Product Delivery
12PUBLIC© 2019 SAP SE or an SAP affiliate company. All rights reserved. ǀ
What did you achieve?
Understanding the structure – Reuse the data
“We have come far in understanding what the data from the Stavå bridge is
telling us about its structural soundness and how this could be applied to
other bridge structures to predict possible failure modes …”
“This information about the structural loading is important to understand the
risks of fatigue.”
Optimize Training – with the help of partners
“By knowing the exact weight of the freight loads passing over the bridge, we
are able to calibrate the digital twin’s finite element model to the bridge so as
to more precisely assess its physical state and level of wear and tear.”
Martin Hasle, head of SAP’s Predictive Engineering Insights product delivery team
Interesting to know:
When an ASKO truck passes the bridge theteam automatically receives the vehicle weight and the time of passing – via GPS tracking and a Geo Fence.
Heidelberg
Trondheim
Austria
Berlin
Travelling Europe:Traffic Analysis in Austria,a Proof of Concept
ASFINAG was founded in 1982 and is a federal agency.
Plan, finance, build, maintain, operate and collect toll for 2,200 kilometres of motorways and expressways.
No subsidies from the national budget - financed exclusively by income from the lorry toll, the toll sticker and the route toll.
Good to Know
Business Case
Can transparency help? Open the data vault – use information in a better way
You have the data: trucks on Austrian motorways are sensored withso-called gantries / toll gates
Traffic and ‚movemment‘ information could become foundation forbilling
Truck Traffic Analysis @ ASFINAG
Strategic Targets
Fewer Traffic Obstruction
Fewer Accidents
Sound Economic Results
Commitment to Sustainability
Outlook
Enhance analysis: time dependent and dependent on the trucktype
Simulation of incidents (e.g. accidents) using geo machinelearning technology
…
DB Size: ~50 GB ( PoC data only ) – geo and business datain one database: SAP HANA
All evaluation done in realtime - is only possible with a HANA in-memory
Allows increased data density and quality
Business Value
Analysis of traffic flows for strategic alignments and long-term measures.
Capacity and performance of the ASFINAG network: network expansion, lane extensions, interchanges
Traffic safety, risk assessments, accident statistics, tunnel safety
Information dissemination in the agency – support better management info, alignment, decision making
What to do with the results?
16INTERNAL© 2019 SAP SE or an SAP affiliate company. All rights reserved. ǀ
Heidelberg
Trondheim
Austria
Berlin
Travelling Europe:An attempt to predict crisis – to respond more proactively
17INTERNAL© 2019 SAP SE or an SAP affiliate company. All rights reserved. ǀ
The Project – Background and Approach
AimAllow the „conflict prevention, crisis management, post-conflict rehabilitation & humanitarian aid“ departmentproactive resource allocation as mean to restrict conflictintensity and optimize resource allocation
Available
Resources
Time Time
Reso
urc
es
Re
so
urc
es
Reactive Proactive
Digital MeanData driven early detection of crisis in order tooptimize intervention
Optimize the use of scarceresources and most efficient andeffective use of limited financialmeans
In crisis relief and assistance is allocated reactively
18INTERNAL© 2019 SAP SE or an SAP affiliate company. All rights reserved. ǀ
What it means for Political Analysts
„We want to create a database that integrates all relevant information to support politicial analysis.“
Political Analysis – Today Political Analysis – Tomorrow
Analysis
Analysis
Data
Preparation
Research
Research
Data Driven Model
the project pursues a purely data driven approach
aim is to confirm and enhance previous findings
quantitative approach enables permament monitoringand automization of analysis
Previous models for early crisis detection
Easy and easy to explain
Foundation: expert knowledge on politics
Not purely data driven
19INTERNAL© 2019 SAP SE or an SAP affiliate company. All rights reserved. ǀ
Architecture and Source
More than 40 open source feedersystems:
World Database
SAP HANA
Events
Indicators
News
GeoData
ACLED: Armed conflict location and event data project
UCDP: Uppsala Conflict Data Program
WDI: World Development Indicators
WGI: World Governance Indicators
GDELT: Global Database of Events, Language and Tone
PDP: Phoenix Data Project
GADM: Global Administrative Areas
SAP UI5
SAP PredictiveAnalytics
SAP Lumira
20INTERNAL© 2019 SAP SE or an SAP affiliate company. All rights reserved. ǀ
QuestionFor which countries the client has to expect a significantrisk of more than 50 deads caused by conflict in at least one of the next 4 years?
Source Reuse Global Conflict Risk Index data / Open Source 24 relevant indicators Categories: Politics, Security, Social, Economics,
Geography
*) more info: http://conflictrisk.jrc.ec.europa.eu/
Use Case 1: Early Detection of Conflicts
Results Reproduction of CGRI indicators Additional indicators to enhance the model Robust and precise prediction model with SAP Predictive Enhancement of GCRI by adding sub state level prediction
21INTERNAL© 2019 SAP SE or an SAP affiliate company. All rights reserved. ǀ
QuestionMonitoring the news flow of all countries is cumbersomeIs it feasible to automate the process – at least partly? Isit possible to alert proactively on disruptive changes ofthe news flow?
Source Harmonized news feeds from open source data
projects sources are CAMEO-coded (Conflict and Mediation
Event Observations) Events classified in classes
Results events per Cameo coded class, state and month
follow statistical distributions news flow of the last 30 days is compared with
historical data Anything out of the ordinary triggers an alert on
potential disruptive change
Lumira Storyboard
Use Case 2: Disruptive Change of News Flows
CAMEO 0233
Name Appeal for Humanitarian Aid
Example Oxfam Canada today called on theworld community to help save tens ofthousands of Afghan civiliansthreatened with starvation
22INTERNAL© 2019 SAP SE or an SAP affiliate company. All rights reserved. ǀ
QuestionHow do the players or groups in the regioninteract?
Source Data sources eg Armed Conflict Location
and Event Data Project provide structures fast and simple analysis of players and their
network with SAP Predictive‘s „SocialNetwork Analysis“
Results The chart was built with SAP Predictive
Analytics Depicted is the interaction of the Al
Shabaab Militia with other players in 2015. Interaction are shown if more than 10
persons were killed
Use Case 3: Interaction Patterns
Examples of Battle ( ACLED ) :
Nigeria: Military forces launch an offensive on Boko Haram positions in Damaturu. A total of 14 soldiers were killed in action during the attack.
23INTERNAL© 2019 SAP SE or an SAP affiliate company. All rights reserved. ǀ
Heidelberg
Trondheim
Austria
Berlin
All scenarios are supported by SAP HANA and SAP technologies like SAP Leonardo.
The incorporation of Analytics and IoT andMachine Learning into business processes– this is what the Intelligent Enterprise stands for.
Next time more!
Any more questions in the meantime?
Meet Ainars:
PUBLIC
Thomas Kunert,
Senior Director Public Services, EMEA N, SAP
Digitization - what does it mean for public sector organizations? Season 3: Smart City and Smart Country Cases