Jump to content

Heath Raftery

  • Content Count

  • Joined

  • Last visited

  • Days Won


Everything posted by Heath Raftery

  1. Good points. The power, as it so often does, comes from combining data sources. Tyre pressure with fuel consumption by route - what is optimal? Tyre temperature with pressure, ambient temperature, vehicle speed and driver identity - who's smoking the tyres? Shock events, location on vehicle and vehicle position, aggregated over many vehicles - pothole detector! But Jason's point is a good one - sometimes you just need to start instrumenting and discover the insights later. I'll give you a recent example - the instrumented pneumatic tube vehicle counters we developed for our Smart Parking system report on various parameters that aren't critical to the core product. One started to fail recently in a way we hadn't seen before. Turns out one of the tubes had lost its springiness. But thanks to our collected data, we now have a unique data signature that alerts us to a pending "springiness" failure! We can even see the clear progression from healthy to point of failure, and combined with traffic counts and ambient temperature records, we can predict time to failure. That's really quite powerful, yet we never imagined it when we were designing it. Last thought... how do they measure vertical load from within the tyre? Maybe it's based on internal tyre height at the bottom of the tyre vs at the top to get a measure of deformation, combined with the tyre pressure to get total load on the rubber? I can't see that being particularly accurate!
  2. That's it - it almost doesn't matter how mature one's technology or technical capability is if it is not well aligned with purchasing expectations. On the density matter, I had in mind a way to measure the sophistication of installations - there's plenty of connected things already, but the great promise of the IoT is that it can be done at significant scale do to technology advancement. So if your garage door talks to your car, that's M2M, but if light switch in the house is remotely monitored, then that's further along the IoT path. Coarse, I know.
  3. There was some discussion about defining maturity at the last meeting of IoTAA Work Stream 6 (Startups). While this example wasn't mentioned, there was a general acknowledgement that it's a difficult task. I'm pleased there also tends to be a general understanding that there is a chicken-and-egg challenge in that the IoT market is quite immature, so proudly claiming one's own maturity is misleading at best. I think you're getting at the same thing - until customers are informed enough to write confident purchase orders, any one claiming a level of maturity is just waving their peacock feathers. With so much still subject to debate, perhaps a framework with more quantifiable aspects might be attractive: dollars saved via the IoT; regions covered by LPWANs; density of things; hours of uptime; number of IoT startups; percentage of active participants in an IoT association. But you might be on to something - while there might be a heap of technical capability ready to come online, the IoT can't be considered mature unless people are actually paying for Smart Parking installations, or autonomous farming equipment, or Smart meters or whatever.
  4. until
    March: Sigfox with Thinxtra Our second meetup for 2017 is on, usual time, usual place: 6:30pm, Thursday, March 2nd, 2017. Stag & Hunter Hotel, Mayfield. Upstairs function room - look for the staircase in the middle of the pub. We're fortunate to welcome Guy Langlois from Thinxtra, to Newcastle for the evening. Thinxtra are the national operators for the Sigfox low-power wireless area network (LPWAN). They've been rolling out coverage at a tremendous pace across New Zealand, Australia and most recently, Hong Kong. Guy will be bringing some exciting news for Newcastle as well as some hot give-aways for the maker-minded. It'll be a two-part presentation: opportunity of LPWAN for IoT & case studies very hands-on presentation on how to use Sigfox devkit As usual, outside the presentations there will be a news recap, plus plenty of opportunity to talk business and tech with like-minded folk from the local area. Feel free to bring up a drink from the bar downstairs or even order a meal from the restaurant, plus there'll be free finger food after the talks! All are welcome, but please RSVP so we can get the catering right.
  5. until
    February: The IoT-enabled Automotive Future Newcastle IoT Pioneers is back! The festive break has been good to us, with lots of exciting developments to announce. The first event of 2017 is on Thursday, February 2nd, 2017, usual time, usual place (6:30pm, upstairs function room at the Stag & Hunter Hotel in Mayfield - look for the staircase in the middle of the pub). We managed last year to touch on just about all the major IoT verticals, except the fascinating world of smart vehicles. I'm very pleased to announce that the group's own resident automotive electronics expert, Norman Ballard of Hummingbird Electronics, has agreed to fill us in this month on some happenings in that space. As usual, outside the presentations there will be a news recap, plus plenty of opportunity to talk business and tech with like-minded folk from the local area. I'm looking forward to meeting a lot of fresh faces from all walks of life this year, so keep inviting your friends and colleagues. Feel free to bring up a drink from the bar downstairs or even order a meal from the restaurant, plus there'll be free finger food after the talks! All are welcome, but please RSVP so we can get the catering right.
  6. The spectrum from a PLC to Raspberry Pi is vast. And a PIC is a microcontroller which is another thing altogether. Typically if your application suits a PLC, it does so for reasons of industrial compatibility, operator familiarity and existing software frameworks. If your application is not heavily reliant on those things, then it could be a contender for the likes of a Raspberry Pi or a PIC based solution. If you're looking at >1000 units/year, and you have the budget for custom development, then look to a PIC (or one of many, many, other microcontrollers of which Geoff lists a few - talk to an elec engineer). If you're looking at a handful of units for a short term pilot or similar, then look to the Raspberry Pi. But don't be too quick to write off the PLCs. Have you looked at the CLICK range? It would be hard to get similar functionality from a non-PLC solution for much cheaper: https://www.automationdirect.com/adc/Overview/Catalog/Programmable_Controllers/CLICK_Series_PLCs_(Stackable_Micro_Brick) Non-PLC solutions are attractive if you have specific requirements and the volumes to justify the development. Non-PLC, embedded solutions are my bread and butter, but even I would caution against thinking that the nifty, nimble, low-power and buzzword compliant world of embedded electronics is a drop in replacement for a PLC.
  7. At one point I swore myself off any engineering project that involved (1) wireless comms, (2) self-sufficient battery power or (3) sleeping microcontrollers. Some how I chose to launch a business right in the triple point that is IoT sensors. After losing a day today dealing with a maddening signal strength issue that popped up out of nowhere, I remember why I made that declaration!
  8. This is a terrific story. I hope you're planning to publish wider (perhaps somewhere on LinkedIn?). The story stands on its own two feet, but I'm especially drawn having marvelled at the Plexus/Green Brain system when trying to find an opening for my startup business. We went 80% the way towards developing a remote compost pile temperature monitor with many aspects in common with Plexus, but ultimately the project didn't go ahead. The path we have taken, that of Smart Parking, went live on Friday and your story rings home. We set out with a very simple goal - count cars as they enter and leave a car park facility to determine the total occupancy level. What we ended up with starts with a pneumatic system with road anchors, rubber tension properties and a tuned pressure threshold. It passes through an analog electrical interface to a sleeping PIC microcontroller with very tight operating limits for maximum battery life. Then it's into the RF world of LoRaWAN with carefully mapped RSSI patterns as well as robust packet redundancy and reconstruction. The data starts as bytes from an ADC and gets wrapped into JSON for processing and distribution in the NodeJS environment, Node-RED. It splits and goes via MQTT to an IoT cloud backend where it is further processed in Go to produce the raw data for consumption via a couple of different HTML webapps. The other channel converts to CSV and goes via TCP to logstash to populate the NoSQL database1 Elasticsearch for searching and graphing with Kibana. Sometimes I stop and wonder if anyone realises what world we've managed to build for ourselves. Perhaps it's inevitable - the customer wants simple and powerful, so behind the scenes an army of engineers from all disciplines build their world ever more powerful and compatible. Before you know it, every little "app" requires the coordination of the latest and greatest from mechanical, industrial, embedded, electrical, telecomms, software and UX engineers. And that's before you even manage to define a product that matches a market, or has great distribution, sales and support. 1. Yes, calling Elasticsearch a NoSQL database is a gross simplification. So be it.
  9. It's almost self-fulling. From my interactions with the world of big data of the last year I've noticed a willingness to sacrifice quality for quantity. It's almost as if instead of thinking about the value of the data at the record level, the approach now is to gloss over details like accuracy, relevance and valid representation. Just turn the firehose on and let the big machines with their magical AI sort it out. So now anyone trying to extract meaning from little data finds that there's nothing of value because the data is crap. Eventually you might get enough crap pilling up so that one pile of crap is a bit bigger than another pile of crap and eureka - the big data machine has found something! I like to temper my enthusiasm for this amazing big data device with the cautions of Cathy O'Neil:
  10. So very true. But your willingness to share makes a huge difference. Many times we have developed a product to meet the expectations set by OEM marketing, using the information provided by the datasheets, with hard won expertise from decades of training, only to discover that we're the guinea pigs finding that the practical limitations fall short of the potential. As much as I love to jump on board the next great thing and develop on the latest technology, I've learnt to stay wary of anything that doesn't have a "community" around it. The power of the community is extraordinary, and it explains (in a self-fulfilling kind of way) the success of the Arduino over the STM32 Discovery, the Raspberry Pi over the VoCore, Linux over BSD. It's natural to dismiss these projects as "hobbyist" (and I do, sometimes) but the fact is it is just too time consuming and risky to discover the limitations on your lonesome, no matter how clever and professional you are. You either become the pioneer, investing months and significant dollars to find you've designed your way into a dead-end and the competition is about to leap-frog you, or you navigate the community forums and figure out the capabilities and gotchas from the collective power of the "hobbyists". Fortunately by the time I began my ZigBee journey, the community had caught up, and I was able to do my research and set my expectations without much investment. On the LPWAN front, of great interest to many IoT practitioners, I've had a foot in both camps. I started by waiting and watching for the community to form, to learn from the scouts at the coalface, suffering the early blows. But I've also discovered that given the focus on European and US markets, and the fact that the Australian market actually has unique technical challenges, that there's value in taking some lead in developing that community. I recently learnt I'm only one of 4 people in Australia with a functioning, Australian regulation compliant, LoRaWAN network made up of components from more than one vendor. Those painful steps into the unknown command a valuable, if fleeting, position in the marketplace.
  11. The Newcastle IoT Pioneers meets again this Thursday, the 1st December. Highly recommended for anyone in this forum that's in the Newcastle/Central Coast/Hunter Valley regions. All the event details are here: https://www.meetup.com/Newcastle-IoT-Pioneers/events/234665531/ Just hit "RSVP" if you'd like to come. Newcastle City Council will take the stage this month to talk about their plans to nurture the IoT community, and to draw on the collective efforts of the local area. Cheers, Heath
  12. That's the kicker for me - I worked for a long time on applications for mesh networks, and it was this gotcha that killed it. ZigBee (and the 6LoWPAN family) promised so much and battery-powered (periodic sleeping) mesh repeaters is actually part of the ZigBee spec. Turns out no manufacturer supports it because the practicalities of synchronising sleeping nodes in a reliable way to perform mesh networking without compromising energy consumption is just not feasible. For now, mesh is relegated to academic projects. If each node contributing to the mesh has to be powered/always on anyway, then you may as well go with a more traditional star network. Which is one reason I find LPWANs so exciting - finally there is a low power, low complexity, wireless comms solution for relatively dumb and sleepy nodes. While we wait for national rollouts, we're installing our own LoRaWAN networks to get the job done. I literally just commissioned one last week, so while it is seriously bleeding edge, with all the inherent cuts and bruises, it's into the realm of here and now.
  13. ...and it's up to us Engineers to sail through the muddy waters of the trough of disillusionment. https://www.linkedin.com/pulse/hype-internet-things-over-jim-wowchuk Warning: the article meanders off into a bunch of vague, obscure aviation details. The meat is in the first part I think.
  14. Cisco have granted Adelaide "Lighthouse City" status. What else do you want? This prestigious accolade is bestowed by Cisco on cities that agree to sign an MoU. In return Cisco agree to sell things to them, if they choose. Some refer to this honour by the outdated terminology of "potential customer". But this crude terminology fails to recognise that there's a Hub involved. Thinxtra rolling out a real IoT network has the unfortunate side-effect of actually making a difference, which is risky.
  15. Very nice. As I commented when I saw this on IoT Australia, I would have done it for $50,000. And made 400% profit.
  16. This is a great program that is sure to entice councils into allowing some much needed infrastructure for IoT applications. I am confused just what Thinxtra are giving away though. The installation is already "free" (Thinxtra always funds installation because they make money on subscribers). The dev kits are a nice though they're not particularly expensive and any serious developer will already have their own rather than relying on a communal kit. The free connectivity is the big ticket item here and I couldn't find any details. What qualifies as a "smart council application"? How long is the period of free connectivity? This is Thinxtra's bread and butter, so I'm sure they have plenty of enticing "your-first-hit-is-free" style offers. That's quite an astute observation. This gap between promise and reality is causing quite some pain for those that actually want to make some stuff work in the field. The LoRaWAN hardware situation is very challenging in Australia. I'm less familiar with the Sigfox situation but I'm sure there are similar gotchas. Does anyone have a bare-truth assessment on this situation today?
  17. until
    http://www.meetup.com/Newcastle-IoT-Pioneers/events/234665531/ The first Thursday of December falls on the 1st, when Newcastle IoT Pioneers will meet again at the cosy Stag and Hunter Hotel in Mayfield. Newcastle City Council have been extremely active in fostering the innovation community in Newcastle and recently took some very exciting steps towards nurturing the burgeoning IoT community. The council sees Newcastle transitioning from a primarily industrial base to one that boasts a broader diversity of economic foundations. Rather than backing any particular horse, the council endeavours to maximise opportunities and provide learning, networking and experimentation platforms, so local innovators can do what they do best. As part of that work the council is developing a Smart City Strategy to guide stakeholders towards a smart and innovative future. I'm honoured to have Nathaniel Bavinton, Smart City Coordinator for the City of Newcastle, booked to fill us in on the very exciting initiatives the council have in store for us in the IoT community.
  18. The Newcastle meetup is alive and well! It formed earlier this year with some advice from the good lads at the Sydney event. I gave this community a plug at the meetup we held last week and will keep the calendar posted. Next meetup is 1st December. http://www.meetup.com/Newcastle-IoT-Pioneers/
  19. Another day, another radical new battery claim. Unfortunately this one too seems to be in the realm of make-believe and misrepresentation. The ultra low power transistor itself seems feasible enough, but single transistors do not typically represent a significant proportion of the current draw in an IoT device. So sure, if all you want to do is power the transistor, then the power draw is low. In fact, the self-discharge of any battery will be many orders of magnitude higher, so the battery will run flat just as quickly if it was powering the transistor or not. But any useful IoT device will need to be doing a lot more - powering sensors, radios, analog circuitry, power and signal conditioning, etc. This probably has applications in devices that are already low power enough to be harvesting energy, where every microwatt counts. Maybe it would mean you could add some extra simple functionality without having to add a battery. But if you already require a battery then ultra low power transistors are unlikely to make any difference. I can just imagine the conversation with the researcher:
  20. until
    The Newcastle IoT Pioneers group meets on the first Thursday of every month. We're looking for hustlers, hackers and hipsters in or around Newcastle, Lake Macquarie or the Hunter, who want to take advantage of IoT. Whether you're ready to show off your project or concept, or just keen to see what the buzz is about, drop in or drop us a line. November's meeting is at the Stag & Hunter Hotel in Mayfield. Pending confirmation this week, the topic this month is "Making solar power intelligent with IoT". All welcome, free to attend, and even some free finger food to follow the presentations. http://www.meetup.com/Newcastle-IoT-Pioneers/events/234665520/
  21. Only just came across this community and am very impressed at the contributions so far! Lost a couple of hours just catching up... but time well spent.

  • Create New...