Book Club
Tales of the City
Urban Design


Travails with the so-called "SmartGuide"
Russell Degnan

A few nights ago I was in the city waiting at the super-stop. No wait, SUUUPER-stop, on the corner of Collins and Swanston St. As well as the extremely useful tram indicators that tell you whether you have time to run and get a jaffle pie before your tram comes, it also has the moderately useful SmartGuide.

Being a bit of a nerd, I couldn't resist having a play, and having done that, and being a sort of uber-nerd, I couldn't resist coming back again to try and break it. And break it I did. The moderately useful thing is that it has a Melway map built in on which you can plan a route and scroll around the city. The useless part is that the route finder is a little bit "indirect" on occasion.

I chose a route to some obscure place in the South-Eastern suburbs. The short story is you need to take the Frankston train to Ormond.

Above is what it gave me. Since the text is hard to read, here it is in text:

Catch Port Melbourne Tram 109 at 9:24pm from Swanston St. to Flinders St., arriving at 9:42pm

Catch Wattle Park Tram 70 at 9:42pm from Spencer and Flinders Sts. to Swanston and Flinders Sts., arriving at 9:55pm

Walk 81 metres to Flinders St. Station. Catch Frankston Train at 9:55pm to Ormond Station, arriving at 10:57pm. Walk 904 metres to your desired destination.

Total walking distance: 996 metres

Brilliant I first thought. But.... no.

The first terrible assumption it seems to make is that you have some idea where you are going. Which is why it lists a tram to take from that platform even though it is going nowhere near where you want. Now, the first rule of Human-Computer Interaction is that users don't know what they are doing and will actively break your seemingly logical system. The programmer who did this should be embarrassed. But it is worse than that, because there are lots of reasons people might be trying to do any route from any point on the system. This is a massive failing.

At any rate, this assumption results in a bizarre and unnecessary tram sequence. From Collins and Swanston to Spencer and Flinders, then back up Flinders. Taking a measly half an hour to do something that could have been trammed up Swanston in 3 minutes, or walked in 2. (Leaving aside the possibility to tram down Elizabeth, or get off at Spencer St. station).

Worse still, it assumes trams run on time perfectly, because the tram times match exactly (and the Port Melbourne was already late), so you would not actually catch the Wattle Park tram unless it was late.

A similar error is then made for the Frankston connection. Apparently we have godlike powers to transport ourselves short distances in zero time. A walk of 81 metres (I love the exactitude of that distance) will take 2-3 minutes. Scheduling exactly is plain stupid. Not that it matters, not only is the time to Ormond massively wrong - 1 hour, 2 minutes! instead of 29 minutes, and just 18 from Richmond because the loop is slow - but the timetable is wrong anyway. The train runs on the quarter hour and every 30 minutes in that period. You could saunter up to Flinders St., catch the 9:45pm and be in Ormond by 10:14pm.

But that's ok, because if you are sucked in to taking the tram, its times are wrong too! The Port Melbourne comes through at 9:35pm. The Wattle Park at 9:50pm.

In short, the SmartGuide is as dumb as dogshit, confusing, wrong, with hopeless heuristics on time, a terrible sense of direction, and stupid assumptions. I pity the tourist who mistakenly uses it to try and go somewhere.

Update: It occurs to me that the times may in fact not refer to when the vehicle arrives, but instead the time from which you have to wait for them. In which case, the end point times are correct, whereas the start times are ridiculously counter-intuitive.

Tales of the City 25th August, 2004 22:47:24   [#] 


Well, there you go....
Russ, love your technical evaluation of the Smartguide. Gave me a giggle.
Lisa  26th August, 2004 00:19:20  

Russ, I think what you'll find with more experimentation is that the smartguide system is still very much so divided along the old M> and Yarra Lines, or partially biased towards the services of the operator of the particular stop you were at. I can't be sure, but if my memory serves me correctly, every one of those displays installed on the old Yarra Superstops is biased to providing Yarra services first rather than M>Tram services. With Swanston street being old M> explains the non-offering of the more direct service in your query.
If it is an inevitable choice of mode or service (such as the train to frankston) the computer will offer it.

The old M> smartguides were always being taken to with sledgehammers so there was seldom chance of them ever confusing anybody.

Finally, Nothing beats the service that can be offered by a qualified, knowledgeable and friendly staff member.
Agent FareEvader  29th August, 2004 12:23:50  

AF, that makes sense. As a reason anyway. As a strategy it is completely asinine. Their lines weren't in competition with each other; one set ran north-south, the other east-west. All they've succeeded in doing is to make their guides useless.

On your last point, you are right, which is why they should replace the goon-squads with conductors again. I suspect this might happen in the runup to the Commonwealth Games, but who knows.
Russ  29th August, 2004 14:27:35  


May 2015
April 2015
March 2015
February 2015
January 2015

Recent Comments

Optimal stop spacing and
      Russ, Tony Morton
Mode Choice and Rational
      Russ, Jason Murphy
The Gastronomic Pub Crawl
      Russ, Andrew
Monday Melbourne: CCLIV,
Monday Melbourne: CCXXXII
      Russ, Andrew
Monday Melbourne: CCXXXII


Scribbling on Bricks
Melbourne 2030 Portal
Melbourne on Transit
Save Our Suburbs
Sustainable Melbourne
Urban Creature


The Next American City
Andrew Blum
Architecture and Morality
Bright Lights Dim Beauty
       of Chicago

City Comforts
The City Desk
City States
Diamond Geezer
Forum for Urban Design
Me, My Life +

Progressive Reactionary
Rebuilding Place in the
       Urban Space

Urban Cartography
Urban Commons
Urban Planning Research

Design and Development

A Daily Dose of

Beyond Brilliance,
       Beyond Stupidity

Brand Avenue
CoolTown Studios
City of Sound
Curbed [LA] [SF]
The Ground Floor
Lebbeus Woods
The Measures Taken
New (Sub)Urbanism
Private Sector
       Development Blog

Reason Commentaries
Richard Green Sit Down Man, You're
       a Bloody Tragedy

Urban Planning Blog
Veritas et Venustas
Wow Flutter

Culture and Theory

2 Blowhards
Abstract Dynamics
Aesthetic Grounds
Anne Galloway
James Howard Kunstler
Junk for Code
Karrie Jacobs
Life Without Buildings
Martin Krieger
Place Space
Rough Theory
The Sesquipedalist
Side Effects
Space and Culture
Strange Harvest


Blog Like You Give a Damn
The Commons Blog
Environmental and
       Urban Economics

Impact Analysis
Jetson Green
Landscape and Urbanism
Muck and Mystery
The Perfect City
Web Urbanist
World Changing


cfsmtb in low
       earth orbit

Live from the Third Rail
Peter Gordon's Blog
Streets Blog
Train Blog
The Transportationist

Non Blogs

Planners Web
Project for Public

New Urbanism
American Planning

Polar inertia

Australian Policy

Liveable Places
Australian Transport
       Discussion Board

Urban Design Forum
Urban Residue
Environmental News
Metropolitan Transport