The making of a train info voice application

From hope…

Seven (!) years ago, a Silicon Valley based co-worker of mine at Tellme Networks developed a speech/phone interface providing local Caltrain railway information, by screen scraping their website. Somewhat later, when Yahoo, Amazon, Google and E-Bay started providing open APIs and thereby led the pack in making the Web a programmable platform, I started wondering how long it would take for this evolution to take root in Europe and Belgium as well. Tim O’Reilly, who coined the term Web 2.0 in September 2005, was in Brussels one year later for EuroOscon, so I went to listen. My frustration levels rose to unhealthy heights when I compared the USA policy on open data access with the European one, which amounts to having taxpayers pay twice, or more for their own public data.

To frustration…

A few months ago Infrabel (”When rail means service”) launched, in a long time overdue effort to bring real-time railway (delay) information into the hands of the daily train users. The lack of a mobile website was promptly filled by the local development community: enter with its iPhone-friendly version. Its makers had to screen-scrape data from the railtime website, as Infrabel does not make its internal web service available to third parties (yet?). Instead of joining forces, Infrabel replied with a quick-fix mobile version on – decide for yourself which is more user-friendly.

It’s also good to know that Infrabel’s SMS version has been announced for at least six months now. And an update of the main Railtime website should see the light by “the end of 2009″.

When I contacted Infrabel about the availability of a web service for entrepreneurial speech technology developers like myself, I got a friendly reply stating that alternative data access methods were “not currently on their agenda” and that “they would contact me again when the need for a voice interface would arise”. Bummer!

What interests me in this discussion is the opinion of the daily railway users … in this year of Darwin, shouldn’t they be the ones to decide on which applications get to survive after they have seen the light? One thing is for sure: in the long run, inbreeding is a bad innovation strategy. So if parental planning of innovations is a must, it’s certainly not up to the government (controlled companies) to distribute condoms, let alone decide who should use them.

And salvation?

You can imagine how my anger turned into joy when (Flemish) government sponsored IBBT announced iMinds and the INCA Awards. Their explicit reference to public data initatives from the UK and the USA gave me that exciting proverbial “now or never” feeling. So I decided to give it a go and developed a voice application in something like two weeks (including weekends, I must say). The prototype provides spoken phone access to real-time Belgian train delay information, as offered by Infrabel’s Railtime website.

So now it’s up to the jury and the public to decide. Not so much if my humble submission deserves an award, but if as a society we keep on accepting the costly inertia of our government authorities and the public companies that they control in terms of open data access.

So let your voices hear, loud and clear!

And may the best win.

Need to write a lab report help? Explore Need help with writing? You have Custom review writing help at Explore best personal statement writing service – Term Papers for Sale – 100% Original Written Paper –