EE Times: So, what was the original vision of wireless sensor network?
Weiss: I wasn't a co-founder of Dust Networks, but I joined the company early on. Considering the expected advancements of sensors, microcontrollers, and radio, our vision then was to create this tiniest thing -- operated by battery or by using energy harvesting -- that can sense, detect, and take the information in the physical world and use it in applications.
EE Times: That's how the idea for a wireless sensor network was born?
Weiss: Bear in mind that there was nothing new about the sensor network. Sensors have been around. A game-changer was the idea to do it at an ultra-low power and without wires. I mean, no wires for both communication and power.
EE Times: What conditions did your wireless sensor network have to meet in order to cater to the real-world applications?
Weiss: First, we needed to develop a large-enough industrial wireless network -- that must be very reliable -- to detect what's going on with their industrial processes.
Second, we had to use no wires at all -- no wires even for power. Unlike other communication technologies such as Bluetooth and cellular phones whose power only last so many hours and they expect to be plugged in for recharge, we had to build a wireless network so that people can walk away from, for example, pipelines in Siberia, not thinking about replacing batteries for many years.
Third, this had to be a wireless sensor network simple to deploy. We're working with people walking around with tools in their tool belts. They weren't going to replace what they've used for decades (communication protocols like HART standard in industrial applications) with IP. The wireless sensor network had to be so simple that it required no training.
EE Times: How do you compare your technology with others?
Weiss: There is a range of technologies that play in a wireless sensor network. It includes everything from a garage door opener to Bluetooth, which is great in a personal network, and Near Field Communications (NFC), which you literally need to be next to a thing that you need to operate. They are all good technologies but they're not very reliable for a wireless sensor network that must covers nodes in a hundreds of acres of physical environment.
EE Times: What was the impact of Linear's acquisition on Dust Networks?
Weiss: In early days, we only had three people trying to sell our wireless sensor network. With Linear's acquisition, we now have a couple of hundred sales people. We're also exposed to a vast number of applications and companies Linear works with in the industrial, automotive and communications market. Linear, by design, is not into the consumer market, and that's well aligned with our focus.
EE Times: Looking back on early 2000s when nobody was talking about IoT, what differences do you see now?
Weiss: Ten years ago, the wireless sensor network was my soapbox. That was the thing all my buddies and I thought the world needed. The wireless sensor network has then become a subset of today's bigger IoT world, where a cast of characters -- like Cisco, General Electric, and Intel -- are trying to sort out a range of issues including the network's efficiencies, operating cost, and big data.
The wireless technology [for both data and power] and high reliability are still very important and relevant [in IoT].
EE Times: Dust Networks today have two product lines: SmartMesh WirelessHart and SmartMesh IP. Do you encounter customers who tell you that they want to transition from WirelessHART sensors to IP-based wireless sensor networks?
Weiss: For legacy systems in industrial applications, it's important to preserve the Hart standard from the asset management perspectives. When Hart sensors in the network interface with an infrastructure, the system needs to know what to do with data.
In a case like SierraNet program, where the system measures temperature, humidity, snow depth, soil moisture, etc., in remote environments, SmartMesh IP-enabled sensors do a great job for researchers working on the web.
Wireless sensor network installed at Duncan Peak in the Sierra Nevada mountains. Sierra*net (http://systems.berkeley.edu/wsn/), an initiative between UC Berkeley and UC Merced, is aimed at studying mountain hydrology and improving the forecasting and control of water supplies by using SmartMesh-enabled sensors.
But if it were a wireless sensor network in oil refineries, no. Our customers value the legacy system that runs on the Hart standard.
EE Times: Of all the technologies Dust Networks pioneered for wireless sensor networks, which specific building blocks are you contributing to the IP-based networks?
Next page: Contribution to 6LowPAN network