Patrick Mannion

's profile
image
Boss

Biography has not been added


Patrick Mannion

's contributions
  • 08.02.2017
  • ZTE kickstarts NSA 5G NR tests: What is it and why should we care?
  • Hi DSPer, I hope this helps make it clearer: Wireless sure has its share of acronyms. So here goes: 3GPP: Third Generation Partnership Project OFDM: Orthogonal frequency division multiplexing LTE: Long Term Evolution TDD: Time division duplex FDD: Frequency division duplex MIMO: Multiple input, multiple output I think I got 'em all, but let me know if you'd like further explanation: happy to help. Thanks!
  • 07.13.2016
  • Despite Tesla accidents, autonomous vehicles are closer than they appear
  • Hey GregB110, I thought of your comment when I was trying to get my fancy car radio to connect properly to my handset over Bluetooth (Mazda CX-5 and Android S5). It wouldn't play media, only phone calls. I put it aside, for the time being, but shook my head in amazement that it is still so hard to get these two systems collaborating effectively, nevermind a fully autonomous vehicle running around Manhattan (I'm a "Nu Yawkah"). I don't think we want machines to be thinking like humans: we have enough humans thinking like humans. We need machines to complement us, or be better a certain tasks, if and when they're up to the task. Unsupervised autonomous driving along public roads, if that ever happens, isn't going to happen soon, though Ford will tell us otherwise. It's good they set a target, kind of like a Kennedy-esque moon shot, and I wonder what we'll learn along the way, even if the ultimate isn't reached, for a long time.
  • 08.24.2016
  • Making fog computing sensors clearly reliable
  • Ah, ok, I see what you're asking. Mesh networking is concerned with ensuring data gets from point a to point b by the most efficient and reliable means possible using a multitude of nodes as relay points. It's not concerned with the content of the data or whether or not to send it along. Fog computing, on the other hand, is the IoT equivalent of distributed intelligence. The originating node (typically a sensor) has enough intelligence on board to decide if the reading is worth sending upsteam (i.e.: if the temperature or vibration is out of normal range), or if it has the capability to control the parameters itself. The goal is to limit the amount of data being sent to the cloud and to decrease latencies (and save power, improve security etc.. etc...) So, it's more intelligent than mesh at each node from the edge to the cloud. About real-time monitoring: Yes, true. But the goal of the monitoring in this instance is more in line with IoT's uber purpose: To gather data on and analyze many -- many -- sensors in situ and track failure modes over time to provide predictive maintenance and also improve product design through the remote, collective analysis. This help?
  • 08.24.2016
  • Making fog computing sensors clearly reliable
  • Hi MWagner_MA, I'd like to answer but first: Are you asking if: A: Fog computing is the same as mesh networking or... B: If real-time monitoring and analysis of sensors for reliability purposes is the same as mesh networking, or.. C: Something else in the story (??) is the same as mesh networking. Help me out here:) Thanks! Help me out:) (And sorry if the writing led to any confusion.)
  • 07.13.2016
  • Despite Tesla accidents, autonomous vehicles are closer than they appear
  • The classic default scenario of fully autonomous vehicles ferrying us around the standard roads we have now is unrealistic, at best. Autonomy will likely come to goods transportation vehicles along predetermined, specially defined and outlined paths. The main point of the article is that long before we get to fully autonomous vehicles we will be in this limbo where we have some autonomous capability, such as on main highways, but we need drivers to still stay fully alert, not dozing off, playing games, texting or otherwise distracted. How we prevent that from happening, as we start to rely more and more on the car doing for us what we should be doing for ourselves, is the design and sensing challenge. And opportunity.
  • 06.22.2016
  • Wake up and listen: Vesper quiescent-sensing MEMS device innovation
  • Hey Steve, I have to disagree with the notion that the best way to communicate is via voice. And it's not because of my accent:) Voice is useful, but when you're in a crowd or quietly communicating online or writing, voice is awkward. It's one way, but I don't see a time when it'll be the "best" or only way. Vesper: Congrats!
  • 02.23.2016
  • GPS-based mood sensing could save lives
  • I agree with all you guys: it's invasive, researchers have too much time, and it could be put to poor or negative uses. It's also laden with use-case flaws (if I forget to bring my S5 phone for my jog, it tells me later in the day to get moving - dumb phone -- but same could be the case here: if a patient forgets or deliberately leaves it in a drawer, it would set off false alarms). That said, it's a prelude to something that may be applicable if modified and implemented correctly. It's very early stages, but may help get to something useful. I did a quick scan and it turns out that 2/3 of the reported 30,000 suicides in the US alone are due to depression (http://www.dbsalliance.org/site/PageServer?pagename=education_statistics_depression) so if you or anyone else can find a use for the research, who knows, maybe it'll raise a flag in time for someone. Or, the researchers could be on the wrong track entirely.
  • 02.09.2016
  • Temperature sensors are improving, but select carefully
  • Quick correction: it's type-J thermocouples that use iron (not type-K). A type K thermocouple uses chromel. But, the point is valid: reserve type-J (iron) thermocouples for inert and low-humidity uses. Maybe I'll lay this out in a short piece , as it's easy to mislabel the types and their metals if you're not immersed in it every day.
  • 01.13.2016
  • 4 reasons Fitbit is being sued for inaccurate heart-rate monitors
  • BTW, a much cheaper way to do this whole HR monitor thing is to count your beats at your wrist for 10 seconds and multiply by 6. If you have a HR monitor. Calibrate it using this method and it should compare nicely and that'll give some assurance as to your device's accuracy.