pgrouting 2.0 presentation 2013

Post on 17-Dec-2014

2.128 Views

Category:

Technology

5 Downloads

Preview:

Click to see full reader

DESCRIPTION

In 2013 pgRouting was presentated at three FOSS4G events, in Nottingham, Tokyo and Osaka. In case you have missed them all, take a look at the presentation slides.

TRANSCRIPT

Shortest Path Search with pgRouting

What is pgRouting ?

An Extension for

PostgreSQL / PostGIS,

...

An Open Source

project, ...

A Library providing, ...

Shortest Path

Driving Distance

Traveling Salesperson

v2.0

Feature Highlights

➔ Two new All Pairs Short Path algorithms

➔ Bi-directional Dijkstra and A-star algorithms

➔ One to many nodes shortest path search

➔ New TSP solver (with distance matrix)

➔ Turn Restricted shortest path (replaces Shooting

Star algorithm)

➔ A collection of useful utility and graph analytics

functions

More Highlights

➔ Modular library design

➔ Unit tests and automated tests with Travis

➔ Compatibility with PostgreSQL 9.1+ / PostGIS 2.0+

➔ Installs as PostgreSQL EXTENSION

➔ Added pgr_ prefix to functions and types

➔ Improved build process for Windows

➔ Better documentation in several languages

https://www.transifex.com/projects/p/pgrouting/

Big THANK YOU to

- Steve Woodbridge

… and the pgRouting community!

- Regina Obe

Most users need pgRouting for

Road Networks

How do they look like?

Like this ...

… or like this ...

… or sometimes like this.

What makes them real?

Traffic lights Signs Road marking

pgRouting extends

PostgreSQL/PostGIS

What does this mean?

SQL Function

SELECT * FROM pgr_dijkstra(' SELECT gid as id, source::integer, target::integer, length::float8 as cost FROM ways', 30, 60, false, false);

Query Result

seq | node | edge | cost -----+------+------+--------------------- 0 | 30 | 53 | 0.0591267653820616 1 | 44 | 52 | 0.0665408320949312 2 | 14 | 15 | 0.0809556879332114 3 | 13 | 14 | 0.072694271986776 4 | 12 | 13 | 0.081239348480584 5 | 11 | 12 | 0.00746935522787469 6 | 10 | 6869 | 0.0164274192597773 7 | 59 | 72 | 0.0109385169537801 8 | 60 | -1 | 0(9 rows)

gid,source,target,cost,reverse_cost,x1, y1,x2, y2,rule, to_cost

source

target

Traffic lights slow down

… so costs must increase.

1 min

cost(A) = cost(A) + 30seccost(B) = cost(B) + 30seccost(C) = cost(C) + 30seccost(D) = cost(D) + 30sec

B C

D

A

Signs inform about

restrictions and rules

cost(A) = length(A)

reverse_cost(A) = ∞

A

Sometimes

the costs

have different

meaning.

A

B

cost(A) = length(A) / 2

reverse_cost(A) = length(A) * 4

Road type can be used

for cost calculation.

Not only road types

but also conditions

And you know

what is great about pgRouting?

All costs are dynamic!

… which is the opposite to pre-calculated

If the road is closed ...

… or there is an accident, ...

… there is a sign with restrictions limited to a certain time, ...

… bad weather conditions ...

You don't need to

rebuild and reload

your network

You only need to adjust the cost

for this particular road,

… and the next search will go another way.

Flexiblity vs. Speed

Cost can be virtually anything

pgRouting can be used for

different kinds of networks

Canals and Rivers

Hiking trails

… or any other kind of networks

Roadmap for Version 2.1

➔ Add “Multi-modal public transit routing”

(GSoC 2011)

➔ Add “Time dependent shortest path” algorithm

(GSoC 2011)

➔ Add “Two Queue” Algorithm (GSoC 2012)

➔ Add DARP and VRP (GSoC 2013)

➔ Add Graph network partitioning (GSoC 2013)

➔ …. further improvements and bug fixes.

https://github.com/pgRouting

Website: pgrouting.org

Documentation: docs.pgrouting.org

Workshop: workshop.pgrouting.org

日本語 NEW workshop.pgrouting.org/ja

Support: pgrouting.org/support.html

More Information

… or talk contact us directly:

➔ Daniel Kastl daniel@georepublic.co.jp

➔ Hal Seki hal@georepublic.co.jp

Photos from sxc.hu and

flickr under Creative Commons Licence.

top related