Matt Lowe is the co-founder of ZeroKey. Focusing on indoor positioning, they build an advanced technology that provides hyper-accurate indoor locationing. Critical for industrial, retail, logistics, and supply chain users, localizing objects in three dimensions with millimeter accuracy is now a reality.
How Is Millimeter Accuracy Even Possible?
The time of flight measurement is implemented based on ultrasound, acoustics, and a proprietary IP. Add high-level signal processing, and you have a far more accurate tool than radio frequency that most indoor positioning techniques rely on.
Triangulation, Multilateration And Time-Of-Flight Measurement
Time-of-flight measurement is the time it takes a signal to travel between two points.
Multilateration is taking measurements that are based on the time of flight and figuring out how far away something is. Objects emit and receive ultrasound wavelengths and as they move through a geographic area. Objects at fixed locations record distances to mobile objects based on time of flight and propagation ultrasound waves. Recording these distances from multiple fixed locations means that a location for mobile objects can be triangulated.
What’s The Hardware Requirement?
Tags or trackers can be as small as a coin or much larger. They get attached to things you need to track and make up a tracking network – wristbands on people working on assembly lines or clip-on devices for access control, for example.
Products and pallets can be tracked with a tag secured with an adhesive back in a supply chain scenario. The tags communicate with the anchors ( fixed Locations ) that act like GPS satellites. Several anchors and tags form a mesh network, and objects are tracked based on their positions to the anchors. The use of ultrasound and acoustics in this way is very similar to how GPS works in our cars and cell phones.
What Does The Mesh Network Look Like?
The mesh is an essential component of the solution. There is no need for a wire to run to every device to communicate to it, and you can still have all of those various types of tags and devices dotted around in your environment. You can get data from one point in the network to another using the nodes in between to relay the information. To create a high density of nodes, they are dotted around every 30–40 meters.
The result is a dynamic network with objects moving around. If it’s a sprawling environment, you can scale your solution up to virtually any size because the mesh network will support you all the way.
Would This Work In A Library With Walls And Shelves?
Absolutely. Ultrasound and acoustics don’t propagate through walls, but this just means that they would need a slightly denser network of nodes.
Imagine a couple of dozen anchors for each section of your library. Acoustic waves propagate around smaller objects, like your shelves. Not unlike a stream where the water passes around the rocks in the middle. That’s how acoustic waves would work in the example of a library, too. However, if you go up the same stream and you build a dam – a wall in your library – the water can’t flow through. This simply means that you need a higher density of network nodes to account for obstructions that block the propagation of ultrasound waves.
How Do You Attach Meaning To Objects?
You need to know what you’re tracking, and that will largely depend on what industry you’re in.
Advanced manufacturing, supply chain, logistics, and health and safety all categorize their objects differently. In all cases, you need to know what each tracker is assigned to, to be clear on what it means. Context is king. Bluetooth and Ultra-Wideband have lower accuracy and positioning fix. As a result, the context may be obscured, especially in high fidelity scenarios.
Imagine an automotive assembly line. You can very specifically digitize the workflow of the workers by providing them with a tracker. If they happen to be attaching a water pump to an engine, you can see if they’ve put it on correctly and if they torqued the bolts in the correct order. You couldn’t really discern these quality measures if you’re using less accurate systems like Bluetooth or Ultra-Wideband. Differentiating torquing a bolt on the left and then on the right that might just be a couple of centimeters apart is now possible. Traditional indoor positioning systems cannot distinguish between the order of the two because they are not accurate enough. But when you have it all down to millimeter accuracy with technology, you can distill the context so that your data becomes more insightful and impactful in the digital solution you’re using.
How Does Hyper-Accurate Indoor Positioning Solve My Problem?
Take a dangerous area in your facility. You need to track when people deviate as much as five centimeters from an authorized pathway because if they do, they put themselves in danger. You can put tags on individuals and alert them through a vibration that they’re off the path. If you only had a meter level of accuracy, you wouldn’t be able to accomplish this. You can avoid fatalities and other serious accidents resulting from workers accidentally entering off-limit areas or not following guidelines correctly. This is just one of the examples where accuracy can be used for something no one had thought of before simply because up until now, it hasn’t been possible to achieve that millimeter accuracy on a large scale.
The possibilities are endless, especially in supply chain scenarios. What you can do with automation is remarkable. Artificial Intelligence, integrating multiple systems, and digitizing physical objects that have been entirely invisible to the digital world will have an enormous impact on what we can do.
In a warehouse, human operators go around and carry out inventory counts and decide for themselves on the most optimal route between tasks assigned to them. If you have digitized people, you’ve digitized the environment. You can now send your employee, Tom, who is closest to aisle 11 to aisle 13 to pick up the next thing because it’s the most efficient thing to send him there and not someone else. Data will let you understand what’s the most optimal and efficient, and you can direct the human resources accordingly vs. relying on the human element to do what they think is the most optimal.
Is This For Humans Or For Machines?
Humans and robots are already working together. Cobots, or your robotic colleagues, are already using data as part of a system or as an instruction to lift a robotic arm, for example. But this isn’t just for the digital world. It’s also for humans, like the ones working in a warehouse. If you equip them and they can carry out real-time quality control, they have a higher Return on investment than the traditional way of implementing a process and then robotizing it.
Local Coordinate System or Earth Coordinate System?
ZeroKey has its own local coordinate system, on a per-network basis which is also relatable to Earth coordinates. The system is smart enough to output GPS or Earth coordinates as well as a local reference to a known point and produce x, y, z coordinates.
To put it in a GIS specific context, let’s look at autonomous drones, which, as we all know, have their own challenges. Their main challenge is doing perfect repeat landings 10,000 times out of 10,000 times come wind, rain, or sensor failure. Even a meter off your supposed landing spot could hit someone or a car. Or in the best-case scenario, crash your drone. GPS (or GNSS) is only accurate to 2-4 meters, on average. Hardly the millimeter level guidance your drone needs to land where it should 100% of the time. In this type of dynamic environment, you need low latency information to control the drone; otherwise, you face havoc on the control systems.
Now you can apply technology to a landing zone situation that’s previously only been used in industrial settings. It’s fusing two technological innovations together. Precise location capability and drone technology. It’s mind-boggling if you think about it.
When Will I Be Able To Walk Into A Building From The Outside With Seamless Interaction?
The market is there for the ubiquitous indoor location solution, but the critical mass is still needed for a scalable technology.
Bluetooth is the closest we’ve got to that critical mass, but it can only address around 70% of the market. It’s not accurate enough to cover the full market. Factor in the high maintenance costs of Bluetooth, and you can see why it’s not mainstream. We are still waiting for global traction to mainstream the technology so that people don’t need a different app every time they walk into a different supermarket or shopping mall. This may also require a significant technological jump for GPS positioning.
Back in the old days, Congress wasn’t convinced that GPS would have civilian benefits. It was a foreign concept at the time, a step too far. It was something we weren’t used to or understood. It wasn’t until the capability was in place that GPS became ubiquitous to the point where you can get a GPS receiver off the shelf. Once it got to smartphones, the use cases exploded, and its economic impact is now off the charts. It’s hard to envision where things can go.
Indoor positioning impacts just about everything in your day to day life. Most of your objects, as far as digital perspective is concerned, are at an unknown location: your house, your car, and your keyboard. Your computer doesn’t know where your keyboard is relative to anything else, neither does your desk and window. They are entirely invisible to the digital world. Turn your camera on, yes, but your things are still not localized. Indoor positioning is about to change all that fundamentally, but it’s hard to see that far ahead. ZeroKey is taking baby steps on this enormous path, but it’s exciting to go and run with this thread.
Can you see a process in your GIS workflow where you’d benefit from hyper-accurate location technology? Did you immediately go into sci-fi mode and imagined robots serving your favorite beverage while you’re sitting in your chair? Or is this too Matrix for you? Let me know what you think.
