Not a member yet? Register for full benefits!

A Sensor Web Built from Every Screw?

Sensor webs are wondrous things; those of us who deal in augmented reality and data mining of all sorts are well aware of that. They allow live, streaming data of practically any type that is desired, from any area of the physical world they are deployed in. Individual points of data blanketing a large area, allow for aggregate data collection in ways that would otherwise be prohibitively expensive or simply impossible.

There are of course many deployment problems with sensor webs. How do you get them to carpet the target area thickly enough to be of any use? How do you guarantee they'll be at key points? How do you ensure they stay powered and running when you truly need them to be?

Well, one German team have a unique answer to all of these questions, in their design of structural stress sensors. Screws, nuts and bolts are integral to man-made structures; they are literally everywhere inside of them. So, why not let the screw be the sensor?

A fully functional sensor screw, current generation.
Credit: Paul Glogowski

The idea does have a few kinks of course. Not least among them is that with present manufacturing methods at least, a sensor screw is much more intricately complex and therefore much more expensive to produce than a normal solid metal screw. However, you don't have to replace every screw with a sensor variant. Rather you only need to cluster small groups of sensors at key points in the building’s structure and you'll still get the same benefit.

So how do they work?

The sensors and the screws start out separate, and are combined using basic metal forming technology. Instead of making solid screws, the forming technique makes hollow screws, into which the sensors are basically just dropped. There was a fair amount of careful design involved to ensure that the screws don't lose structural strength from their hollow cores, and of course the hollow is tapered so that the insertion can only go one way; once inside, the sensor unit does not come out again.

The sensor unit is of of course a motion detector, able to detect slight stress-related changes and shifts around it, so it is essential that it is firmly held within its' host screw. To this end rather ironically, it is screwed tight into the larger host screw with two very tiny screws.

In addition to the sensor, a power source and a transmitter are required. Whilst the developers were unwilling to give details of the power generator the sensors use, the fact that it only needs to sound the alarm when stresses cause shifting in the structure around it, combined with the fact that the screws are often going to be located in entirely inaccessible places by intent, narrows the power source down to only one possibility: A piezoelectric source.

In other words, much like how a passive RFID chip gains the power it needs to function from the reader that is broadcasting a radio frequency towards the tag, the sensor screw derives the power it needs to send an alert signal that the structure is shifting, from the very movement of the structure itself. A piezoelectric generator of this type, and in the tiny space available, will only be capable of generating a few millivolts at a time; but that is literally all such a tiny sensor requires.

Like with other such sensor installations, they will only have to transmit a signal a handful of feet at most; a receiver unit will be wall mounted somewhere nearby a sensor group to pick up the readings. The receiver unit having a far more standard power source.

It makes much more sense for the screws to be doing the detecting rather than a wall unit of course, because the screws are by their very nature implanted right at the points of greatest stress inside the structure.

They don't have to be implanted in buildings directly either. One example of a great use for these screws, is to use them to assemble a working machine in a factory or mining environment. Heavy machinery is always vibrating when in use, but if the vibrations become excessive, that is often an early warning that something is starting to go wrong. Sensor screws in key places would be able to detect the increase in vibration, and in addition narrow down which part of the mechanism the discordant vibration is emanating from, by as simple a process as triangulation.

They don't need any real computing power themselves to triangulate with either. As with all sensor webs, the sensor's job is to detect and report. The unit it talks to, deals with the heavy duty computation. This keeps the power costs and complexity of the sensors as low as possible.

Other uses of the screws become obvious at this point. Used in aircraft they can warn of imminent failure of a part of the vehicle long before it actually fails and the plane falls out of the sky. Used in prosthetics, they could warn of stress-related failures to implanted or otherwise inaccessible parts of the unit. The uses start to boarder on the truly ubiquitous. Such a sensor design could easily be used in every machine and structure we have – depending of course on the extent to which the sensor itself can be shrunk.

The Developers

The development team themselves, are German researchers, from Technische Universität Darmstadt
in Hessen, Germany. Developed as a side-spur out of another, commercially driven project to monitor and control the level of uncertainty in operation of any load bearing mechanical system, the screws were initially intended as a way of tracking how the load is affecting the performance of said system by detecting the stress on every joint in real-time.

Thankfully, researchers Manuel Ludwig and Matthias Brenneis realised the potential of what they had designed, and that it had more than enough merit to stand on its own as an independent project. In early 2013, development of the screws split from being part of the original project, and the researchers have secured enough funding for 18 months of further refining. By that point they hope to be producing them commercially and state they already have a number of commercial partners.

From left to right: Manuel Ludwig, Matthias Brenneis, and an early prototype of the sensor screw.
Credit: Paul Glogowski


Their primary concern at the moment, is making it smaller. Currently the screws they can make are solely in hexagon bolt designs, as in nut and bolt. Self-tapping screws are considerably smaller, and a goal is to integrate the sensors into those too, as well as researching refinements to the overall design.

But, that said, make no mistake, this is not an early prototype or proof of concept. The screw sensors work, and are viable now. The researchers simply believe they can make them better than their current design. The German government, in the form of the German Federal Ministry of Economics and Technology has expressed their full support for and backing of the sensor screws.

Whatever happens now, it is for all intents and purposes certain that these screw-based sensor webs will start cropping up in EU buildings, machinery and vehicles, adding an ability to monitor their behaviour within tolerance like has never before been possible.

References & Further Reading

The sixth sense in mechanical engineering

Collaborative Research Centre SFB 805 – Control of uncertainty in load-carrying mechanical systems
The original parent project that spawned the sensor-screws.

The Institute for Production Engineering and Forming Machines
The part of Technische Universität Darmstadt where the researchers are based, and from whence the first screws were made.

Technische Universität Darmstadt

Sensorische Verbindungselemente durch Kaltmassivumformung (Sensory fasteners by cold forging)
The original paper. German only (Subscription Required)

Staff Comments


Untitled Document .