Sensing Paper: A Revolutionary Approach
MOOS leverages 15+y of research to turn paper into sensors
Unlocking the Promise of Connected Shelves
Connected shelves have been promised for decades. During the peak of the IoT boom, approximately 10 years ago, promises were made left and right that 'soon, every shelf will be connected'. Consultants fueled this promise with connected device projections and cases, demonstrating the benefits of having shelves capable of communicating transactions or inventory. These benefits include avoiding stock-outs, optimizing operations, communicating promotions, or creating broader systems for shopper engagement, price optimization, or an unmanned agenda (scan & go or grab & go).
While nobody disputes the potential benefits, getting the technology right seems challenging. Despite several technological advancements in connectivity, energy, and sensing techniques contributing to overall feasibility, traditional approaches seem flawed.
- Cameras and image processing encounter processing and privacy issues.
- Tagging products (NFC, RFID) is often prohibitively expensive in operations.
- Traditional mechanical solutions, like a weight scale, could measure fairly precisely but have a fixed form factor, limiting flexibility, and are intrusive and cumbersome in operation.
- Printed matrix pressure sensors, which are high-tech, expensive, and offer limited response range.
A simpler, more flexible, and far more cost-effective solution is needed to unlock wider application. What if connecting shelves were as easy as placing a sheet on top of it and configuring and recognizing products easily? This was the original ambition of MOOS and led us to go back to the drawing board. Making a better scale, camera, or sensor is clearly not the answer. A rethink of the underlying principles of connecting shelves was needed to break away from traditional technology and vendors.
Introducing Sensing Paper: A Revolutionary Solution
It all started with the notion that a 'signal' from a shelf does not need to be perfect or very precise to capture the full value of the insight. If you could see that you're 'running low' or an event, like 'an item was picked', you can achieve nearly all of the benefits. Exact precision suddenly becomes a nice-to-have, not a need-to-have.
This created an opportunity to use materials that are inherently flexible and cost-effective but come with some noise and usability issues. And... we found it, in one of the most universal materials in business: paper.
The idea is that carbon-loaded or conductive paper changes the resistance with force, creating a pressure sensor if you let current flow through the surface and place a product on it. This principle can be leveraged to create a differentiated sensor.
Sensing Paper: How Does It Work?
We obviously can't just use any paper, not even any carbon-loaded paper. Turning paper into sensors places some special criteria on the paper and the paper-making process.
First of all, the type of carbon, additives, processing steps, and target characteristics are not known. You need to get these right, and ideally, understood and controlled in such a way that the paper gives a controlled response to force.
Traditional paper manufacturing focuses on optimizing their process for characteristics like grammage, strength, consistency, color, smoothness, and functional specs, like hydrostatic properties. However, how to create good electrical or resistive properties is typically not known or measured. Let alone, the 'sandwich' effect of paper, i.e. ability to compress and decompress, creating a natural dynamic range in the resistive response in the Z-axis, while limiting the resistive flow to adjacent areas in the XY-axis. The latter is important because you don't want to short-circuit or create a binary, on/off, circuit. The paper, the carbon particles, and fibers need to work together in just the right way to create compression and associated resistance that can be used to create a sensor.
At MOOS, we are very proud to have teamed up with Prof. Koehly, who has laid the academic groundwork for this in his 15+ years of research into technical paper. Together, we have taken his theory and lab experiments and turned this into the first production run ever for sensing paper.
The MOOS paper is at the heart of the paper-based sensor. A sealing, electronics, and protective enclosure are added for practical operations. We will continue to optimize the properties for a sensor going forward, especially with the interplay with a software layer to interpret signals. Already, at this stage, we have created a powerful, differentiated path to create connected shelves.