The 2010 Trend Watch Sensor Survey results were published by Design News and half of the survey respondents design products for the industrial market. The balance of respondents works in industries as diverse as automotive and aerospace to packaging and healthcare.

When selecting a sensor, respondents said that reliability, accuracy and durability/ruggedness are the top three characteristics, while product support and availability are critical when selecting a particular supplier.

The file with full results is available here [pdf]

Optimizing battery life and long-term reliability involves numerous variables, including the chemistry, the cell design, the quality of mechanical components, the purity of raw materials and the manufacturing processes employed. Shortcuts in quality can negatively impact service life.

The total amount of active chemical ingredients and the ratio of each ingredient determine the cell’s nominal capacity. Predicting expected operating life solely on the cell’s nominal capacity can be misleading however, as the cell’s capacity is affected by the active components, the internal self-discharge, the application power profile and environmental factors. Since the volume of active ingredients is limited by the size of the cell, nominal capacity values often do not vary substantially. So the key differentiator often involves the inner structure of the cell and the ratio of active ingredients. For this reason, design engineers should evaluate the battery’s Equivalent Operating Capacity (EOC) to properly calculate its expected operating life, taking into account the cell’s self-discharge rate, application current profile and environmental conditions.

More info here.

Several years ago, I watched Vint Cerf, who helped draft the architecture of the Internet and is now chief Internet evangelist at Google, give a talk about the future of the Internet.

During his presentation, he discussed the early days of the Internet, when he was developing the protocol called TCP/IP with the United States Department of Defense. He talked about some of the strange early networking experiments his team did, but he also talked about his socks. He explained that one day everything would be connected to the Internet, including his socks, and if one should fall behind the washing machine while he was doing laundry, it would be able to notify the other sock of its whereabouts.

The basis for this concept is called “the Internet of things.”

The day when we have communicative socks might not be too far off, according to a report released Monday by McKinsey & Company. The paper highlights some of the major changes that will result from the growing ubiquity from sensors and objects connected to the Internet, including “sensor-driven decision analytics” and “complex autonomous systems.”

The complete article is available here.

In challenging environments, the required link budget necessary to maintain a given Packet Error Rate (PER) can be quite significant, which is not compatible with the constrains introduced by inexpensive radio architectures and low-power requirements which define WSNs.

RadiaLE, recently presented at EWSN 2010, is an open benchmarking testbed that allows performance evaluation of Link Quality Estimators (LQEs), it aims for experimentation with existing and future LQE implementations.

More info here

Collections of autonomous smart wireless sensors that could connect together to form ad hoc networks were not practical until recently. But in the late 1990s it became clear that Moore’s Law would eventually take performance up and power consumption and form factor down, to the point where they would be possible.

And the juxtaposition of sensors (and potentially actuators), conversion circuits, digital logic, power and power management and wireless transceivers is a potent combination because it is, in essence, the superset template for a vast swathe of electronic systems. So not only were wireless sensor networks becoming possible, they were set to be revolutionary.

And that is where the fun begins because myriad questions quickly follow. What are the right applications for wireless sensor networks? What is the right architecture for the right application? What about the trade-offs of performance against range against power consumption against cost? Is energy harvesting viable to avoid battery costs? What is the right radio protocol? What legacy hardware, software and protocol stacks can be reused? What are the right standards to ensure interoperability?

More info here.

As it relates to wireless sensors, RF energy is the only controllable, practical technology to provide power over distance to multiple sensors simultaneously.  Other technologies are either too directional for one-to-many powering (i.e. IR LEDs), or have severe range limitations (i.e. induction, MR).  There are the critics that say RF power is not efficient and most of the energy is wasted.  However, using RF to power sensors at long range (e.g. energy management and building automation) is not about the efficiency of the charging mechanism, it’s about enabling applications and achieving greater system-wide efficiency.  Having a transmitter than consumes a few watts but provides power to sensors which feedback data to control thousands (or tens of thousands) of watts or BTUs provides a significant “energy ROI”.

Results are as follows:

Academic Master Thesis Award: Matthias Wilhelm for his work: “Implementation and Analysis of a Key Generation Protocol for Wireless Sensor Network“.

Industry Master Thesis Award: Carlo Alberto Boano for his Thesis: “Application Support Design for Wireless Sensor Networks“.

Academic PhD Thesis Award: Olga Saukh for his Thesis: “Efficient Algorithms for Structuring Wireless Sensor Networks“.

Industrial PhD Thesis Award: Antidio Viguria for his work: “Market-based distributed task allocation methodologies applied to multi-robot exploration”

Our warm congratulations to all awardees!

More info with photos at the ceremony could be found here.

The tests performed in the Monegros Desert (Spain) had the purpose of testing the capabilities and limits of the 802.15.4/ZigBee radios integrated in Waspmote. These results were presented at the European Wireless Sensor Network Conference (EWSN2010) last week in Coimbra (Portugal) as a collaboration between the Royal Institute of Technology – KTH, Stockholm (Sweden), the University of Cape Town, Cape Town (South Africa) and Libelium.

Among the 6 different links (356m, 639m, 1239m, 3810m, 6363m,12136m) were chosen Line of Sight (LOS) and Non Line of Sight (NLOS) configurations which were tested always using omnidirectional antennas (2dBi, 5dBi).

Read the complete article here.

More info here.