TfL has been working on broader integration projects where we focus to get the most efficient use of our road networks and public transport. We bring together a wide range of data from multiple disconnected systems which we not only use for operational purposes, but also make more of them open and available; in real time. This session presents how we brought together IoT and Big Data techniques to understand current and predicted transport network status and plan to evolve the base solution into a broader product.
3. Breadth of Transport for London
§ 30 million journeys daily
§ In addition to all road and rail
transport, we look after rivers, assisted
travel, taxi and private hire regulation
§ We do much more also from education
to running the world’s largest out of
home advertising estate of its type
§ We’re 150 years old and chock full of
heritage and design assets
4. Road Space Management
§ Sponsorship
- Own the improvement strategy for the TLRN
- Engage with a wide range of customers and external stakeholders
- Deliver user benefits that are clearly defined and measured
§ Outcome Delivery
- highways design and engineering
- traffic modelling capability
- intelligent control of traffic signals
- Monitor, analyse and optimise road network performance
§ Operations
- 24/7 Operation to keep London moving
- Real-time incident management through LSTCC
- Assess and coordinate works and schemes to minimise disruption on TfL’s roads
5. Wikipedia Definition
§ Digital transformation is the change associated with the application
of digitaltechnology in all aspects of human society. ...
§ The transformation stage means that digital usages inherently enable new types
of innovation and creativity in a particular domain, rather than simply enhance and
support the traditional methods.
6. The proposition
§ Understand the impact of cloud native
§ How do we support portals and data mashups
§ How is the breadth of data incorporated in a relevant display
§ How flexible and future proof
§ CAPEX v OPEX
7. Conceptual Architecture
Key Characteristics
§ Bringing information and insight
closer to decisions
- Locale, map based UI
- Timely, event driven not batch
- Trusted, consistent and accurate
§ Public cloud hosted
§ Reusable commodity platforms
§ Open Standards
Integration
Data Hub
Collaboration
and Innovation
Analytics and
Visualisation
Data Driven
Operational
Applications
Secure Enclave
Data Driven
Operational
Applications
Secure Enclave
8. Data Hub
Commodity Cloud
§ Scalable Object Store S3
§ Persistence Services (SQL and No SQL) Aurora +Dynamo
§ Elastic Search
§ Spark
§ Compute EC2
Configured Platform Services
§ Spatial - ESRI
§ Data Warehouse - Oracle RDBMS +Redshift
§ SOA Platform – WS02
Data Hub
9. Public Interfaces
§We currently provide a number of (digital)
road information sets including:
- Incidents - Live traffic information and
planned works/events disruption.
Includes planned closures.
- Road status tool – Web interface
- VMS (including some journey times),
JamCams, Post Code impact areas and
Other Camera locations.
Many users, applications and value added
providers
10. Analytics and Visualisation
§ Everything we monitor is linked in time and place
§ Improvement in identifying effects and finding cause
§ Balance of models and sensors
§ Where you have multiple data sources how do you bring these
together to inform you of new insight
§ How feasible is prediction ?
11.
12.
13.
14.
15.
16. Analytics Summary
§ Looking into how we present data and how trusted is it,
differentiating
- Modelled
- Measured low density sensors
- Measured high density sensors
§ Bringing multiple dimensions together coherently is the key feature
§ Planning and Operational co-ordination activities need and use the
same information
§ How to introduce real time analytics
17. Operational Systems
§ TfL has an operations center which coordinates on the day
activities across Surface
§ Many events are planned, some events just occur
§ Plan à Do à Review
§ Situational Awareness is important
§ Silo systems just do not work
18. Situational Awareness
§ Should be common across operations
§ The ability to understand what is happening
§ The ability to identify where things are unusual - quickly
§ Sensor coverage is seen as limiting factor
§ Waze data is being trialed
19. Situational Awareness
§ Trial of Waze crowdsourced has
just started
§ Early results are promising for
incident detection
§ Solution combines several data
sources to enrich real time view
§ Management views and
dashboard included
21. Traffic Control Sensors
§ We have about 14,000 sensors measuring junctions approaching junctions
§ This data is currently used by the Real Time Control to manage optimisation
§ We have been investigating how it can be processed for insight
- Scale 780 Million events per day
- Latency to data center circa 1 Second
- Resolution 250mS scans
26. Key points
§ Sensor data is great for modern analytics platforms
§ Meta data preparation is the hardest part
§ Its completely anonymous
§ Granularity of data gives early identification of incidents
§ ML predictions out to 45 mins, are quickly running at >80% accuracy
§ Complex network, understanding its dynamics needs a lot of work
27. Operation Technology
§ Traffic Management Act Notifications (TMAN) - A dedicated interface between
London boroughs and TfL enabling the balanced delivery of major schemes and
works on the TLRN and SRN
§ Forward Planning Tool - An advance planning tool that allows promoters to
provide early visibility of road and street works
London Works 2
§ Central Register – a pan London system enabling
visibility and management of works and related activities in
London
31. Review
§Use Technology as a commodity
- Open Source is a key enabler
- IoT, Big Data, Cloud don’t resonate
- Skills are the biggest hurdle to overcome
§Bringing information sets together encourages new thinking
§Agile changes the organisation
§Composable systems delivers solutions quickly