Post by thread:[EE] Wireless Sensors

This is more of a marketing (ugh) question as opposed to technical... There is so much talk about wireless sensors in the industry publications, but I don't know anyone who knows anyone who uses these. There are quite a few companies (start-ups, most likely) that are offering different transceiver options that do routing and mesh networking and all sorts of fancy stuff or just simple point-to-point. The DOD is looking for "smart dust" to be deployed in war zones. Berkeley has their (open-source, I believe) "motes" with custom software that is being marketed and developed in parallel by another company (Crossbow Technologies). Does anyone on the list have any knowledge or experience of these things leaving the theoretical domain to be used in somewhat large scale? I'm thinking along the lines of building automation where the danger of having wireless sensors unavailable due to interference isn't a real big deal and people might be interested in the technology to reduce installation costs. The use of wireless sensors in laboratory settings seems more of a gimmick than something with very much practical use in the near future; although, I expect that will change in time. For example, Kele has some stuff on their website: http://www.kele.com/cgi-bin/navigate.cgi?section=Network%20and%20Wireless I would also suppose that oil well operations might be leaning towards wireless due to scale of the plants and the harsh conditions that sensor cables might otherwise have to contend with. Any thoughts on these matters or parallel matters? Thanks Bradley

> This is more of a marketing (ugh) question as opposed to technical... > > There is so much talk about wireless sensors in the industry > publications, but I don't know anyone who knows anyone who uses these. I don't know about wireless sensors, but I thought the same way about wireless computer networking a few years ago. > Any thoughts on these matters or parallel matters? Well, I can think of a few places where wireless sensors would make sense: 1. Retrofitting sensors into an existing building without having to rewire the place. 2. Anyplace where you need to move sensors around very often. -- D. Jay Newman ! Polititions and civilations come and spam_OUTjayTakeThisOuTsprucegrove.com ! go but the engineers and machinists http://enerd.ws/robots/ ! make progress

Look at ZigBee (http://www.zigbee.org ?) Chips Maunufactured by Freescale, Atmel, Chipcon & a few others. Microchip support for the Chipcon version using an 18 series micro. RP On 02/06/05, Bradley Ferguson <.....bradleyeeKILLspam@spam@gmail.com> wrote: {Quote hidden}> This is more of a marketing (ugh) question as opposed to technical... > > There is so much talk about wireless sensors in the industry > publications, but I don't know anyone who knows anyone who uses these. > > There are quite a few companies (start-ups, most likely) that are > offering different transceiver options that do routing and mesh > networking and all sorts of fancy stuff or just simple point-to-point. > The DOD is looking for "smart dust" to be deployed in war zones. > Berkeley has their (open-source, I believe) "motes" with custom > software that is being marketed and developed in parallel by another > company (Crossbow Technologies). > > Does anyone on the list have any knowledge or experience of these > things leaving the theoretical domain to be used in somewhat large > scale? > > I'm thinking along the lines of building automation where the danger > of having wireless sensors unavailable due to interference isn't a > real big deal and people might be interested in the technology to > reduce installation costs. The use of wireless sensors in laboratory > settings seems more of a gimmick than something with very much > practical use in the near future; although, I expect that will change > in time. > > For example, Kele has some stuff on their website: > www.kele.com/cgi-bin/navigate.cgi?section=Network%20and%20Wireless > > I would also suppose that oil well operations might be leaning towards > wireless due to scale of the plants and the harsh conditions that > sensor cables might otherwise have to contend with. > > Any thoughts on these matters or parallel matters? > > Thanks > Bradley > > -

In the industrial automation sector, cabling is one of the large portion of the cost. Therefore comes the idea of fieldbus. Instead of one to one cable connection (each sensor to PLC/DCS), you will have multi-sensors connected to a bus (eg. ASi bus -- actuator sensor interface), then through a gateway (eg: ASi to Profibus gateway), all the sensors are connected to the PLC/DCS. Now comes the wireless, you cut the last "mile" of cable from the system. The problem with Zigbee is still the with current consumption and cost. Adding 10s of mA to a sensor which only consume less than 100mA is a lot. Adding US$5 to a US$25 sensor is as well quite a lot. The space is another issue even though PCB antenna is getting smaller. Another obstacle is that this sector is quite conservative. The PLC/DCS makers control the market (Siemens, ABB, Honeywell, Rockwell, Invensys, Yokagawa, Omron, AB, etc). Chipcon seems to be the leader now. Parts from Freescale are now very expensive. Regards, Xiaofan -----Original Message----- From: Richard Prosser [rhprosserKILLspamgmail.com] Sent: Thursday, June 02, 2005 7:16 AM To: Microcontroller discussion list - Public. Subject: Re: [EE] Wireless Sensors Look at ZigBee (http://www.zigbee.org ?) Chips Maunufactured by Freescale, Atmel, Chipcon & a few others. Microchip support for the Chipcon version using an 18 series micro. RP

remote sensing/managing, is SCADA system. very expensive and powerful. with SCADA sit there in high price, every one have to be expensive. ----- Original Message ----- From: "Chen Xiao Fan" <.....xiaofanKILLspam.....sg.pepperl-fuchs.com> To: "'Microcontroller discussion list - Public.'" <EraseMEpiclistspam_OUTTakeThisOuTmit.edu> Sent: Wednesday, June 01, 2005 8:53 PM Subject: RE: [EE] Wireless Sensors {Quote hidden}> In the industrial automation sector, cabling is one of the > large portion of the cost. Therefore comes the idea of fieldbus. > Instead of one to one cable connection (each sensor to PLC/DCS), > you will have multi-sensors connected to a bus (eg. ASi bus -- > actuator sensor interface), then through a gateway (eg: ASi > to Profibus gateway), all the sensors are connected to the > PLC/DCS. > > Now comes the wireless, you cut the last "mile" of cable from > the system. The problem with Zigbee is still the with current > consumption and cost. Adding 10s of mA to a sensor which only > consume less than 100mA is a lot. Adding US$5 to a US$25 sensor > is as well quite a lot. The space is another issue even though > PCB antenna is getting smaller. Another obstacle is that this > sector is quite conservative. The PLC/DCS makers control the > market (Siemens, ABB, Honeywell, Rockwell, Invensys, Yokagawa, > Omron, AB, etc). > > Chipcon seems to be the leader now. Parts from Freescale are > now very expensive. > > Regards, > Xiaofan > > -----Original Message----- > From: Richard Prosser [rhprosserspam_OUTgmail.com] > Sent: Thursday, June 02, 2005 7:16 AM > To: Microcontroller discussion list - Public. > Subject: Re: [EE] Wireless Sensors > > > Look at ZigBee (http://www.zigbee.org ?) > Chips Maunufactured by Freescale, Atmel, Chipcon & a few others. > Microchip support for the Chipcon version using an 18 series micro. > > RP > --

yes -we have a case where installation cost is equal to hardware cost. By going to radio we could knock this back to less than 20%!. The current consumption issue can be reduced by putting things to sleep for a while - depending on application. The Nordic chips also seem to be a good compromise between performance & cost, but are a single sourced solutiion, as opposed to ZigBee, which is at least "supposed" to conform to a standard. RP On 02/06/05, Chen Xiao Fan <@spam@xiaofanKILLspamsg.pepperl-fuchs.com> wrote: {Quote hidden}> In the industrial automation sector, cabling is one of the > large portion of the cost. Therefore comes the idea of fieldbus. > Instead of one to one cable connection (each sensor to PLC/DCS), > you will have multi-sensors connected to a bus (eg. ASi bus -- > actuator sensor interface), then through a gateway (eg: ASi > to Profibus gateway), all the sensors are connected to the > PLC/DCS. > > Now comes the wireless, you cut the last "mile" of cable from > the system. The problem with Zigbee is still the with current > consumption and cost. Adding 10s of mA to a sensor which only > consume less than 100mA is a lot. Adding US$5 to a US$25 sensor > is as well quite a lot. The space is another issue even though > PCB antenna is getting smaller. Another obstacle is that this > sector is quite conservative. The PLC/DCS makers control the > market (Siemens, ABB, Honeywell, Rockwell, Invensys, Yokagawa, > Omron, AB, etc). > > Chipcon seems to be the leader now. Parts from Freescale are > now very expensive. > > Regards, > Xiaofan >

> In the industrial automation sector, cabling is one of the > large portion of the cost. Cost of installation, not operation and total cost over the life-span of the installation cabling cost can be insignificant. > Therefore comes the idea of fieldbus. Any bus is a nice idea, but you also need to identify determinism. EU Class 4 industrial automation requires deterministic control which pretty much eliminates "fly-by-wire" for safety controls, unless you look at some pretty high-end solutions, which negate the cost of running several 110 or 24 volt wires. > consumption and cost. Adding 10s of mA to a sensor which only > consume less than 100mA is a lot. Adding US$5 to a US$25 sensor > is as well quite a lot. The space is another issue even though Adding $5 to the cost of a $20k+ industrial project is nothing if it eliminates service problems and increases the uptime of the equipement. __________________________________ Discover Yahoo! Have fun online with music videos, cool games, IM and more. Check it out! http://discover.yahoo.com/online.html

On 6/1/05, D. Jay Newman <KILLspamjayKILLspamsprucegrove.com> wrote: > > This is more of a marketing (ugh) question as opposed to technical... > > > > There is so much talk about wireless sensors in the industry > > publications, but I don't know anyone who knows anyone who uses these. > > I don't know about wireless sensors, but I thought the same way about > wireless computer networking a few years ago. I think from this comment stems the root of the issue with wireless in industrial settings. Wireless networking is successful (as far as I can see) as a consumer product. Coffee shops, et al. offer wireless as a value-added product. Lay-people use wireless because they don't want to string wires around. The increase in the use of laptops as they've improved has helped wireless networking. Wireless sensors, on the other hand, are more the domain of the professional engineer, who, if they are good, should not automatically specify the newest glitzy technology. They will first look to use what actually works and then they will try to improve it. Wireless has to offer a lot to get them to move away from the reliability of wires. This well-founded trust in the technology doesn't come into play in the consumer market. I don't think wireless sensors are going to have the same growth curve as wireless networking because of this. The question is, then, what's going to spur adoption? > > Any thoughts on these matters or parallel matters? > > Well, I can think of a few places where wireless sensors would > make sense: > > 1. Retrofitting sensors into an existing building without having to > rewire the place. > > 2. Anyplace where you need to move sensors around very often. Yes, this seems to be what the limited market is. You don't have to drill holes in finished walls. I can't really think of moving sensors very much except on very large machines or rotating machinery. Both these seems like such limited markets, that they won't be able to spur a real boom. > D. Jay Newman ! Polititions and civilations come and > RemoveMEjayTakeThisOuTsprucegrove.com ! go but the engineers and machinists > http://enerd.ws/robots/ ! make progress Politicians... Looked weird, I had to look it up. Thanks Bradley

On 6/1/05, Richard Prosser <spamBeGonerhprosserspamBeGonegmail.com> wrote: > Look at ZigBee (http://www.zigbee.org ?) > Chips Maunufactured by Freescale, Atmel, Chipcon & a few others. > Microchip support for the Chipcon version using an 18 series micro. That's part of the reason I bring this up. It seems every industry magazine I receive has ZigBee somewhere on the cover. I've not found any real (not vaporware) products that are actually being installed yet. As another poster mentioned the industry is very conservative and I'm wondering where the inroads are (if there are any) and where this technology is /really/ going. I had forgotten that Microchip has a ZigBee demonstration board. I might buy a couple of those. Bradley