July 16, 2008 – Using sensors over a mesh network, Marius Gafen, Senior Marketing Director of Israeli company Virtual Extension, relates how UIDs can save money and natural resources.

I understand that your company has been involved with certain applications using mesh networks in utility servicing. Can you tell us the work you have done with water meter usage?

GAFEN: We make wireless networks for sensors, for connecting sensors and for metering. There are three types of metering for utilities: electricity, water and gas. We are involved in all these types of metering.

Metering, or as they are called today, AMI, Advance Metering Infrastructure, means that instead of a person coming and writing down the meter reading, this reading is sent by wireless to the central location of the company.

In the application in Jerusalem for water meters, my understanding is you are able to monitor the meters through this wireless network without a meter reader having to appear. Can you explain that with a little bit more detail?

GAFEN: The meters are of a special type, having a smaller electronic circuit, which counts the process for each cubic meter or whatever the flow of the water is. This reading is sent through the network to a sensor location. So instead of a person coming over and noting the reading of the meter, everything is monitored at the base location.

There are some addition advantages. For example, a simple PC circuit can be set to check if there is a leak somewhere, Water is very precious in Jerusalem, so a loss of water is important. If there is such a loss of water, it is important to monitor that. Similarly, we can monitor and sound an alarm when there is a waste of water.

I would guess you are monitoring an unusual amount of usage. You are not actually sensing that there is wetness, right?

GAFEN: No, no. The way that the meter appears there is a rotating spin, which is magnetically monitored, and the readings are sent to the sensor locations through the network. At the network, at the sensor locations, there is a PC program written by a third party, our partner, who is doing the water management (we are only doing the network).

In addition to reading, they are checking each of the water meters at high time resolution on a daily basis so everyone can know in which period of time there was special water consumption. If consumption deviates, they know it.

Where are the nodes located that the meters are reporting to? They are reporting wirelessly to the nodes in a mesh network, is that correct?

GAFEN: Each node, which is actually an electronic box, is connected to four such water meters. From each meter, there is a wire going into the node. And in each node, there is a piece of electronics which can drive the monitor for the water meters.

For a mesh network, the nodes have to transfer data from one to the other somehow to get to the central database. So, if my home and three others are connected to one node, how does my data go through the mesh network? How does my data transfer to the central headquarters?

GAFEN: Actually, in the mesh network each node serves not only the transmitter, but also there is a receiver and transceiver. It receives the data from the neighborhood, and it transmits it over to the other nodes.

In a regular, a standard type of network, there is a routing. It would direct how the information arrives from one node to the sensor location.

In our approach, and in our technology, any node sends the information to any other node and this way it has additional advantage of more range, more robustness, and lower cost.

So, if one node is out, it is not going to affect the transmission of the rest of the network?

GAFEN: Absolutely. What happens is if one node is out, is then all the promulgations go through all the other nodes and they cover for the specific one.

I believe that you also have an installation that has helped public utility lighting users, reducing the amount of servicing costs of the lighting. Can you tell us a little bit about that?

GAFEN: Sure. The application uses one network to control the lighting, to turn on and off certain lamps. This saves energy because you do not need all the lights to be turned on and off on the whole street. It can be done on the basis of a particular lamp, so it is according to needs, to the hours of the day, to the amount of cloudiness, and so on.. It also monitors the condition of the lamp, itself, saving to the operator money because lamps can be replaced with optimal timing.

What were they doing before that, just replacing that every so many days or so many months?

GAFEN: That is exactly so. Usually there is average length of life for each lamp. Let’s say 3,000 hours. So, what the operator would do based on his experience with the plate or the lamps, would be to replace them at say at 2,500 hours, to be on the safe side. In this case, the smart way of monitoring it, it can be done until the last hour of the lamp and we are saving a large amount of money. These lamps are expensive.