Archive for category Stuff
Making use of University Research
While not the best news article I’ve ever seen The Australian IT have a story about Innovation funding in the next budget.
An interview with the Minister for Innovation, Industry, Science and Research:
The lack of collaboration between business and industry — on which new jobs and economic opportunities depended — was “one of the great weaknesses” of the innovation system, he told a higher education conference in Sydney.
Australia ranked 16th out of 28 OECD countries for research collaboration between industry and universities in 1997, he said.
“By 2004 we had fallen 10 places to be last out of 26 countries,” he said.
I’ve essentially been working somewhere in the void between University and Industry for the last 4 years. On one side of the void there is the Universities. I’ve seen many potential projects from IMPCA which simply end with a research article. And I dare say this is exactly how Universities should operate. They should research concepts on the “cutting edge” of science and make the results publicly available (very public. i.e. give away this knowledge — don’t scrooge the ideas with patents).
Then there’s the other side of the void – Industry. It seems to me that companies are only interested in university research if it is “ready for market”. They simply want to make money instantly out of the idea without wasting time and money “getting it ready”
So, how do you fill this “void” between Universities and Industry? I’m really not too sure! 🙂
I don’t think simply pumping money into some form of “commercial ready” grant works, because you have to fund the right projects which will make money for industry. I suppose the only way to work out which are the right projects is more collaboration between industry and universities.
But how do you force more collaboration between industry and universities without just offering grants?
Cane Toads Invade Western Australia
It would appear the inevitable has happened.
Quarantine officers found a male cane toad at the checkpoint between the Northern Territory and Western Australia border on Friday night.
The officers reported hearing the calling of the toads at the border late last week and say there is evidence to suggest the pest has established a breeding ground in the East Kimberley at Lake Argyle.
Why is this a problem? It’s a problem because cane toads kill anything which eats it. Eggs and tadpoles kill the little things (i.e. frogs and geckos), toadlets kill the medium sized things (i.e. lizards) and adults kill big things (i.e. crocodiles). So they wipe out the native entire carnivorous population. And we don’t want a country full of herbivores!
What is being done to prevent this? I think the general attitude is best summed up by the term “meh”.
There are a few community based toad squashing groups (who catch and kill individual toads). I suppose they have the required enthusiasm (hell, they’re living in it, while I’m sitting in front of a desk in Perth), but I don’t think they will affect the situation at all. Official government responses to these groups seems to be a gentle pat-on-the-head.
Ok, How do we stop them? Look at the past. Introduced rabbits were a massive problem for Australia in the early 20th century. Hunting individual rabbits is almost laughable. The big ass fence cost a lot of money but didn’t work. Biological methods appear to have cut rabbit numbers dramatically since mid 20th century. Research and introduction of Myxomatosis and Calicivirus dropped numbers from 600 million to 100 million.
But that was after Rabbits did a lot of damage and money wasted on failed methods.
Why isn’t this proven path being taken to slow the toads? Well, this is where the “meh” factor comes into it. The West Australian government has provided a token $500k grant to begin research as well as funding a report into the “effectiveness of current strategies”. Federally, CSIRO have been researching solutions since 1991, but don’t seem to be getting anywhere in a hurry.
So, for now the only solution seems to be either raise cane toad research to a much higher level of importance (i.e. more money), or just give up an let them take over everything.
Adding drives to a Dell MD1000 storage array
We recently bought a Dell MD1000 storage array (to go with our Dell PowerEdge 2900) with 5TB of space. We’re running Ubuntu (I prefer plain old Debian…. but oh well…).
They say these things are expandable up to 45TB (all on a single virtual disk)… I wanted to make sure it’s really possible! Here’s how I did it:
I ended up using the Dell Open Manage command line interface as I couldn’t get the OpenManage web interface working with Ubuntu. I presume there’s something similar with the Windows install of OpenManager.
See this post to install Dell OSMA and open the manual for the omreport and omconfig commands.
- display the current controllers:
omreport storage controller
- display the current physical disk status:
omreport storage pdisk controller=1
- display the current virtual disk status:
omreport storage vdisk
- insert the new drive (and take note of the bay number)
- reconfigure the virtual disk with the new drive
omconfig storage vdisk action=reconfigure controller=1 vdisk=0 raid=r5 adisk=0:0:9,0:0:10,0:0:11,0:0:12,0:0:13,0:0:14
where 0:0:9 was the new drive and 0:0:10,0:0:11,0:0:12,0:0:13,0:0:14 were the existing ones. - check if it’s finished reconstructing with
omreport storage vdisk
- wait….. wait….. go home…. come back….
- resize the partition and file system in your operating system.For me, this involved using LVM tools to resize the partition, then resize2fs to resize the ext3 filesystem
Google Latitude + GoogleCar
Everyone has heard about Latitude by now. “See your friends’ locations and status messages and share yours with them.”
Ok, there’s a mountain of privacy issues surrounding it… Simple solution — turn it off. I have no intention to broadcast my location 24 hours a day. That’s just asking for trouble. But it could be very handy when you actually want your location to be known.
Anyway – what I’m interested in is how it finds your location?
Most smart phones have some sort of Location API – Here’s an example using the Symbian Location Acquisition API. Java and iPhone have similar API calls.
The other side of the API presumably uses whatever it can to get a location. GPS (possibly with the help of an A-GPS server) or some form of mobile network based “triangulation” seem to be popular.
But what if your phone doesn’t have a GPS receiver? Or you’re using the Latitude iGoogle app on your laptop? Or a desktop?
This is where things get interesting. The Latitude iGoogle app gets your location through the Google Gears GeolocationAPI which according to google, uses “your computer’s WiFi location”. What the hell is your “WiFi location”?
Well, if you can build up a big enough database of WiFi access point MAC addresses and associate each with a location, you can perform a “reverse” query to associate an AP to a location! viola! Each AP should have a 100m (or so) transmission radius, so if there are multiple APs you can “overlay” these transmission circles to get an even more accurate position.
But this theory only works if you have a lot of entries in your database and it’s regularly updated (as people add/remove/move their personal APs). How do you do this? My ideas:
- The Google street view car scans for WiFi signals as it drives around
- Every Laitude program updates the Google database while it is being used
The Street view car visits every street which gives Google a good starting point for the database. As soon as possible after this they get users to “give Latitude a try” which corrects the database with any changes since the Google car went past, making it more accurate as more users come on. To back up this Google car collection claim, I’m going to use a quote from the official google blog “Today, Latitude is available in 27 countries, and we hope to add more soon.”. How many countries do you think Google has Street View imagery for or is currently capturing? About 27 perhaps?
It’s all very clever if you ask me.
Just don’t tell anyone what Virtual Observer is currently doing with rubbish trucks…… 😉
Design Out Crime
Today’s interesting task was attending the Curtin Uni Design Out Crime award presentation. The competition is run by the Curtin design and urban planning bunch and the Office of Crime Prevention of WA with the intention of promoting interesting designs to reduce crime.
The winner was Park Friend – an outdoor seat which doubles up as a bike rack. The idea here was that people are always around the bikes, hence the baddies are less inclined to steal them.
Tele and I came second with Tagless, otherwise known as Graffiti Detect. Our project doesn’t really fit in with the whole “design out crime” thing… but it’s meant to detect window scratching on buses, trains, shops etc through a piezo microphone and some signal processing.
Third was the No Climb Bin – explains itself really.
The most interesting discussion here was how to prevent all this behavior in the first place rather than just catching people as they do their naughty stuff.
Wilson catches up with the Joneses
Well, I’ve finally decided to check out this whole “blog” thing.
It all happened after talking to a Brett Adams from IMPCA. He tells me that a lot of people do this blogging stuff now and I tend to believe Brett, because he knows a lot more about stuff than I do.
Now It’s time to see if *I* can bore the world with my ramblings! Hoorarh!
Expect to see posts about whatever I’ve stumbled across during my morning coffee. This might include stuff related to:
- My work: Virtual Observer
- Research in the Computer Science field
- Happenings in the area I live in
- and maybe even a few political rants if I feel up to it.
Also prepare to be slightly bored. I can’t imagine why anyone but me would bother to read this 🙂
Recent Comments