SANTA CLARA, Calif. The Department of Homeland Security is taking the wrong approach to funding sensor research for anti-terrorism, said the program manager of the SensorNet program at Oak Ridge National Laboratory.
Moreover, the industry needs to come together around a handful of key standards to accelerate the development of sensor networks, said John Strand in a keynote address here.
"The DHS is spending [funds] on sensor development, but not on sensor networks and interoperability," said Strand in an interview with EE Times following a talk at the first IEEE conference on sensor and ad-hoc networks here.
"They haven't looked at an overall architecture for how to put the pieces together. As a telecom guy that scares me. We could wind up cobbling systems together from pieces in a way that's inefficient," he added.
The two-year-old SensorNet program has deployed about 200 sensors in three test beds as part of an effort to establish a national sensor network for first responders to crises such as terrorist attacks. The Defense Department has bankrolled much of the SensorNet effort. The program has a budget of about $20 million this fiscal year expanding to about $30 million next year.
In his keynote address, Strand said industry standards for various hardware and software interfaces are needed to accelerate the rise of sensor networks. "To do this at the national level, standards need to be involved," Strand said.
"One of the things we found out is manufacturers aren't using the IEEE 1451 interface [for linking sensors and controllers], so we have develop wrappers for sensor data. That's been one of the most demanding things we have had to do," Strand said, noting his group has developed about 15 sensor interfaces.
However, some experts at IEEE Secon (Sensor and Ad-Hoc Communications Networks) said the 1451 standard is geared more to the needs of wired system-level data loggers, not the emerging generation of wireless sensor nodes that integrate sensors, controllers and radios on modules the size of a deck of playing cards.
"It's true 1451 carries too much weight for tiny sensors," Strand admitted. "Its an open question whether we should have an interface at the sensor or the node level," he added.
The SensorNet group has launched an effort to define requirements for an applications programming interface for sensor networks used in embedded control systems. It expects to have a prototype API in about nine months.
"The money in sensor networks will ultimately be at the applications level," Strand said.
He noted the University of Alabama is already working on a variant of the XML language for sensor networks.
Another focus area for the SensorNet group is security and privacy.
"We are looking into the development of a policy server that could be managed at the local level and would require a separate security server. Most of the data from sensor networks won't be highly classified but it will be protected for privacy," he said.