Sensing as a Service

Matthew WoottonBlog

Benefits of Occupancy Sensing

Let’s start with occupancy sensing as a baseline for making the environments we live in truly smart. If you don’t know something has happened, you can’t do anything about it. To date, the world has thrown all kinds of hardware and bespoke sensing solutions at the problem, with each solution aimed at a narrow set of use cases. Unless there’s some miracle coming, hardware won’t be able to address the breadth of use cases the market is looking for today.

The fact that occupancy sensing to date has relied upon hardware means there is a lot of hesitancy to deploying or testing different opportunities on what could be done if we knew what was happening in an environment. Occupancy sensing is the right decision, especially for smart spaces, but nobody wants to buy and maintain hardware specifically for those purposes.

With hardware being difficult and expensive, software based sensing solutions are the best path forward. Occupancy sensing in all spaces is what’s needed in the end, and sensing needs to be tailored for what it’s used for. Being able to dynamically update and evolve sensing over time is something that hardware just can’t do. Hardware has to be managed, maintained, replaced, etc. Software is simpler to deploy, maintain, and manage. The world of occupancy sensing is still awaiting that killer use case. Unleashing sensing services on the world will unlock a wide range of new opportunities and capabilities for systems that we’re not even able to imagine today. Enter Sensing as a Service…

Sensing as a Service

Ok, now let’s get into Sensing as a Service. Stemming from the fact that more sensors allows for more intelligence, it stands to reason that the best sensor isn’t necessarily the most capable, but rather the most broadly applicable. Add to that the notion that any device which either senses or reacts to sensing must have a communication chip and you build a recipe where the world is full of devices for network presence sensing deployment.

Human nature is one of ratcheting experiences. Once the smartphone was here, we couldn’t imagine the world before it, even though folks got by just fine. The smartphone became a necessity, a necessity based on the software solutions which were added on top. If not for those essential software additions, I might still be using a flip phone today.

Why did that software matter to its adoption so much? It made the hardware more useful, but more importantly it lets a developer give experiences to millions of users at the press of a button, and also allows the collection of data on a scale that was unforeseen to that point.

That same opportunity exists today with Sensing as a Service. The breadth and depth of deployment that can be achieved with software can really show users what they’re missing. Many folks that add PIR to their systems end up loving it, despite the many drawbacks of such a sensing solution. Rather than make users buy another device before they know how good sensing can be, why not show them through a service on the devices they already own.

Now that’s the user side, when we look at the data side and the meaning of that data, that’s really where the world has turned to software for solutions. Want to figure out what someone is thinking, go to social media. Want to figure out what they’re buying, go to credit cards. Want to figure out when someone is home or not and/or if there are intruders, apply sensing as a service. We know that insurance providers and utilities need data to tell them something is really valuable to them. Doing that with hardware just doesn’t work.

In short, sensing as a service let’s folks see what sensing is all about, and think about how they want to use it. You can imagine an occupancy data app store, connecting different use cases to different available data, with the most commonly available data being created through sensing as a service.

These data that are created through sensing as a service would be able to be linked with various applications, and have performance matched to those use cases. Now, the software has to be ready to deploy on as many devices and protocols as possible, not just wifi chips with CSI and a high bandwidth connection to the cloud for compute. Network Presence Sensing lets sensing as a service be deployed locally on almost any platform, without lots of requirements for our partners. In summary, the inherent flexibility of software combined with its ease of deployment really makes sensing as a service the next logical frontier for IoT.

Why does this matter

Looking at why Sensing as a Service matters, we’ve all been waiting for those truly intelligent buildings that just respond to us and we don’t have to interact with them. The only way that works is starting with basic occupancy sensing capabilities. Without an ability to know if a person is there, a building and all of its systems cannot do anything in response to that person being there. On the simple end of the spectrum, walking into the room and having the lights turn on is nice. On the more complex end, you could also leave the house and have all of the following happen: the security system arms itself, the thermostat goes into a deep set state, all unnecessary loads in the home are turned off, and the WiFi prevents new devices from joining. In both uses, they start with knowing whether or not someone is home.

Furthermore, when looking at different use cases, they happen at different timescales and with different relative accuracy requirements for both knowing there is someone there and knowing when there isn’t someone there. As an example, a lighting use case relies on always detecting someone entering a space and doing so rather quickly (less than 1 second), but does not have as stringent a set of requirements around extra detections. A security system on the other hand must always pick up when someone enters a space, and extremely infrequently when someone is not in the space, but can take up to 15 seconds to make such a determination. These data have very different performance requirements in reality; however, most systems treat occupancy data as interchangeable even if it really isn’t.

With Sensing as a Service, data can be constructed according to what it is meant to be used for. You don’t try to find a happy middle ground between the use cases, you simply create separate sets of data which allow each use case to run. Without software and service based sensing solutions, you cannot create bespoke data for different use cases, and deploy additional use cases to existing systems as needed.

What’s more, Sensing as a Service significantly reduces the risk associated with deploying the technology. There’s no hardware changes, simply update the devices to include technology like Network Presence Sensing and you’ll have deployed a sensing service. This service allows existing customers to experience what it’s like to have occupancy sensing installed without going through new hardware installation and lets them try out different use cases with devices they already have. As many of our customers have commented, “it’s like magic.”

Last, but certainly not least, is the ability to run large scale trials of different offerings, collecting massive amounts of data with relatively little expenditure. If an insurance company wants data on a million homes with Network Presence Sensing vs a million homes without it, all that needs to be done after an initial deployment with a hardware partner is piping the data to the correct location. If a utility company wants energy savings data, the same logic may be applied. Leveraging large and previously installed hardware bases, Sensing as a Service allows the creation and understanding of occupancy data with flexibility and scale that has never before been seen. It’s an exciting time to be in occupancy sensing, with the dawn of Sensing as a Service unlocking a whole new level of value for intelligent buildings, the likes of which the world has never seen.