Automation Founders Circle
Micro-robot insects hatch ‘Smart Dust’ node for ISA Sperry Award winner Pister
EDITOR’S NOTE: ISA continues its tradition of honoring leaders throughout the automation industry by presenting the Automation Founders Circle awards. This year’s recipients are Dr. John Ziegert with the Arnold O. Beckman Founder Award; Kristofer S. J. Pister with the Albert F. Sperry Founder Award; Leonard Moore with the ISA Honorary Member award, the highest honor bestowed by the society, and Peter Martin with ISA’s 2009 Life Achievement Award. This month we write about Ziegert and Pister, and last month we featured Moore and Martin.
By Jim Strothman
From the time he was a graduate electrical engineering student in the late 1980s until today, Dust Networks co-founder and Chief Technology Officer Kristofer S. J. Pister, has been fascinated with making the tiniest of wireless electronic devices.
The idea of what later was to be called “Smart Dust” initially dawned on Pister in 1992, when the then UCLA assistant professor of electrical engineering—who at the time was trying to make millimeter-scale micro-robot insects—attended a RAND Corp. military workshop.
“We were looking at the ‘battlefield of the future,’ he recalled, “and I realized I had (in micro-robots) technology tools needed to help with that.
“You need six things in a robot—sensing, computation, communication, power, motors, and mechanisms,” said Pister, who earned his Ph.D. in electrical engineering at UC Berkeley. He currently serves as a professor of Electrical Engineering and Computer Sciences at UC Berkeley and is co-director of the Berkeley Sensor & Actuator Center.
Wingless body was wireless sensor
“I realized if I scratched off the last two things on that list, motors and mechanisms (which for a micro-robot insect includes wings and legs), it became much less expensive. It became a wireless sensor node,” he said.
“Smart Dust was basically the body of my silicon insect,” said Pister, who continues to also consult for the U.S. Department of Defense’s JASON “think tank” several weeks during the summer. Pister first coined the term “Smart Dust” in 1996 and commercialized it in 2002 with the co-founding of Dust Networks, Hayward, Calif.
Pister and others at Dust Networks developed the Time Synchronized Mesh Protocol (TSMP), wireless HART-based technology, using the Smart Dust Sensors as network nodes. TSMP is said to deliver the highest reliability and lowest power consumption of any protocol developed to date in industry or academia.
Named ISA Sperry Award winner
For his technical contributions to “advances in the development of high-reliability, lower-power ‘mesh’ networking, Pister will receive ISA’s 2009 Albert F. Sperry Founder Award.
The honor, to be presented during the ISA Honors and Awards Gala on 5 October, the day before the 2009 ISA EXPO 6-8 October in Houston, Tex., recognizes “an outstanding technical, educational, or philosophical contribution to the science and technology of instrumentation, systems, and automation.”
The award was named for Albert E. Sperry, who was internationally recognized for his contributions to the advancement and development of instrumentation as an innovator, business executive and ISA leader. Sperry served as the first ISA President in 1946 and was elected an Honorary Member of ISA in 1956.
“I was thrilled to get the award,” said Pister, who has played a key role in the ISA SP100.11a wireless technology standards effort. TSMP has been donated to ISA and is a foundational building block for the ISA100 standard and also is the foundational building block for the WirelessHART protocol.
Mesh network ‘paths’
In a TSMP, full-mesh network, each wireless node has routing capabilities—a significant advantage over networks with nodes that do not have the same functioning capabilities or rely on special-purpose routers, base stations, or aggregators. Dust Networks’ chip in each node is about the size of a nickel, and the chip is packaged in products about as small as a deck of cards.
TSMP contains five key components:
Automated node joining and network formation
Fully redundant mesh routing
Secure message transfer
TSMP sidesteps RF interferers (multiple radio frequencies commonly found in manufacturing plants) by “frequency hopping,” referred to as Frequency Hopping Spread Spectrum. Pister said the distance between each node in a mesh network can be “something like 100 meters per hop—there’s lots of data to support that.”
Current TSMP installations operate in the 2.4 gigahertz (GHz) ISM band on IEEE 802.15.4 radios and in the 900 megahertz (MHz) ISM band on proprietary radios. Dust Networks said over 1,000 TSMP nodes can operate in the same radio space with each other, without affecting end-to-end reliability. By comparison, dense networks of nodes using collision-based protocols like Carrier Sense Multiple Access can experience cascading collisions and network failure.
TSMP installation costs and the cost of acquiring sensor data have been shown to result in considerable savings—as much as 90%--compared to other networking technologies. Up to a 20% reduction in energy costs also are claimed.
Traditional wireless has been point-to-point, which can be adversely affected by RF interference, paths blocked by physical changes, and loss of individual nodes. If the points are not within line-of-sight, a distance as short as 50 meters cannot be guaranteed with a point-to-point system. However, with a TSMP network having multiple sensors—each having routing capabilities—a node can turn to another “friend” node to re-direct the data if something blocks the original data path.
“There are literally dozens of choices of paths,” the Dust Networks co-founder said. This ability to self-configure, in effect, gives TSMP networks “self-healing” capabilities.
And, since nodes on the outside edges of a mesh networks have routing capabilities, networks can be easily expanded at minimal cost.
Pister and others foresee applications almost too numerous to imagine—beyond typical wireless sensor network applications including industrial process automation, commercial building climate control, and security alarming.
In process industries, TSMP networks can move data from sensors gathering real-time information about the health of equipment; process data such as pressure, temperature, level, and flow; as well as environmental measurements.
Pister sees a big market opportunity measuring vibration in smaller motors. “If a manufacturer has a 100 horsepower (HP) motor, hooking up vibration sensors is routine. But one- or two-horsepower pumps traditionally don’t get monitored for vibration because it’s too expensive. For Dust Networks, condition-based measurements is a huge market because manufacturers want to know whether their equipment is healthy” before a breakdown occurs. Most plants have many small motors, he noted.
Another “big market opportunity,” he said, are instruments in remote locations, too far away to be monitored via traditional wiring, and checked infrequently because of labor costs and distance issues. By connecting the remote instruments to inexpensive nodes in a TSMP network, plant operators can monitor them as often as they wish.
Works where ‘wired’ does not
Then there are places in plants where wires simply cannot be hooked up. Pister cites a Pittsburgh steel mill where operators needed to make temperature measurements in an environment where red-hot slag, metal, and steam were the norm.
Wireless technology also was an enabler taking measurements on a rotating kiln sloshing in a paper mill, he said.
Wireless also typically works better than wires on equipment that moves around—forklifts, or industrial palates, for example, he said.
For the food industry, what is known as the “cold chain”—monitoring temperature conditions as food moves through the supply chain—is critically important. One of Dust Networks’ venture capital sources is Cargill, a privately owned international supplier of food products, which has considerable interest in that technology.
A November 2008 InTech article (http://www.isa.org/intech/20081104) written by Pister and Greg LaFramboise, Wireless Technology lead at Chevron Energy Company in Richmond, Calif., details Chevron’s wireless installation applications in harsh environments such as oil fields and refineries. It also addresses security, which “was of paramount importance to Chevron,” the duo wrote.
Sees health care applications
Asked to cite potential applications outside of industry, Pister immediately suggested health care, specifically monitoring infants, the elderly, or people who are ill.
“I have four children, and I often found myself popping into their bedrooms at night just to make sure they were breathing and they were OK,” he said.
“With our technology, you could clip a coin-sized thing onto an infant’s pajamas and monitor their breathing and heartbeats, for example, from another room.”
Elderly people could be supplied with bracelets to alert someone if they stopped moving, he said. “That might be OK in the bedroom, but if they are in their kitchen, no.”
Others have talked about mesh network technology being used to monitor for forest fires—signaling alerts before many trees burn, and doing the same for earthquakes or mine cave-ins.
“It’s a long list,” agreed Pister.
In addition to his academic and research achievements, Pister has written numerous technical papers, is a co-holder of several patents, participated as a speaker and session chair at many technical conferences, and has won several other technical awards.