Steve Toteta, new kid from Dust Networks, is talking about an economic perspective on wireless mesh networking.
His agenda:
Wireless Sensor Networking
Economic Impact
Case Studies
Self organizing and self healing mesh technology brings many things to the industrial environment. I don't believe, he said, that we're in the wild west of radio technology. That was in the early 1900s. From 1900 to 2000 there was over a trillion times improvement in spectral efficiency. He also showed a graph of decreasing wireless system cost from Marconi to WiMax. Now we're entering into a very interesting area, that of WSN: wireless sensing network technology.
The goals are very simple:
Achieve more efficient operations
Maximize Plant reliability
maintain safe work environments
detect process and equipment issues
RF is Unpredictable in space, in time and in frequency. Application requirements change. Low Power, highly reliable networks required. This means we need to move from star to star with repeaters to hybrid networks, and then on to full mesh networking.
RF and application conditions are dynamic. The intelligence is in the network, so let the network do the work to optimize.
With star topologies, you typically need experts to help you tune the network, which can become quite expensive.
With a full mesh network, you are relying on the embedded intelligence in the network to reduce the unpredictability of the network.
Site surveys are static pictures. The wireless environment changes. We should not have to depend on site surveys for control of our wireless networks. We need more dynamic control of the network. Frequency hopping radios are what SP100 and WirelessHART are based on, because the system itself can make decisions about optimization.
Battery life must be LONG and PREDICTABLE. Devices must last years in the field. Batteries can be difficult to replace.
Synchronization greatly simplifies the challenge--highly accurate charge accounting. The nodes all know when to wake up and listen. We can tell you with each transaction how many micro coulombs that takes and we can tell you how much battery you have left. In real time you are optimizing the battery life of the network.
Now for the economic impact:
At the end of the day, it is the business imperative that determines whether you implement this technology.
Economic return = lifecycle benefits - lifecycle costs.
Cost of the system
+initial engineering costs
+installation costs
+annual operation costs
+annual maintenance costs
-------------------------
= lifecycle costs
Higher productivity
+annual manufacturing cost savings
+improved safety and reliability
+ improved energy efficiency
+lower labor costs
_______________________
= lifecycle benefits.
The wired Paradigm has 80% of costs in system installation and engineering costs.
Wireless chagnes everything.
Wired costs have a curve that increases over time.
Taking the wires out of it produces some cost saving, but the curve continues to go up over time. If you need to have experts come in and tune the system repeatedly, you can incur some significant continuing costs.
With mesh networking, your costs are even lower for installation and engineering to start with, but also lower over time.
When you pull this all together, you begin to see that technology is an enabler. Toteda predicts a 10x reduction in cost of getting information, distributed intelligence, and ubiquitous sensing.
Now, then, about that case study...
He quoted Dave Lafferty of BP as saying that wireless makes a step change reduction in installed costs. They saved 700 feet of conduit, 3000 feet of wire, etc. in a simple 5 node pilot facility.
