RTLS Studio 2.3 Significantly Boosts Performance and Increases Flexibility
By Lubomir Mraz | August 13th, 2020 | 6 min read
With faster performance, increased flexibility, optimizations for large-scale enterprise projects and improved usability, Sewio RTLS 2.3 is the best version yet, bringing precise, scalable and reliable indoor location tracking to the enterprise world! The Sewio R&D team has put a lot of work and sweat into the latest version of RTLS Studio, which is now available for Sewio customers and partners via Sewio Portal.
- Support for Multi-core Processors: Up to 10x Better Performance
- TCP Sync gen2 Release: Leveraging Machine Learning Genetic Algorithm for Improved Flexibility
- RTLS Studio Docker Container: Introducing an OS-agnostic Solution
- Sensmap Improvements: Better Usability
- Sectional Initialization: Streaming Large Deployments
- Anchor Firmware Updating from RTLS Studio: Faster and Safer Upgrades
1. Support for Multi-core Processors: Up to 10x Better Performance
Nearly all modern computers have at least two CPU cores, with many offering four or more, and RTLS Studio 2.3 is now able to exploit this hardware capability to greatly shorten location calculations, thereby, being able to track a higher number of objects simultaneously.
Let the numbers speak for themselves as I share the results of our evaluation test using the huge positioning traffic created by a high density of anchors.
- 256 Anchors deployed within 16 rooms, a single master per room – test area around 6500 m2
- Tags blink with an update interval of 0.3 s, typically used for tracking medium speed objects. Each tag had visibility to 16 anchors around
- The host server was an Intel(R) Xeon(R) Platinum 8168 CPU @ 2.70 GHz, 16 cores and 32 threads, hosted in the Azure cloud. A native RTLS Studio installation
We have used a single API client, simultaneously gathering positions from both WebSocket and UDP API connectors and checked the number of positions calculated and latency per each API connector.
We are rather conservative, and only show the results that passed the following criteria:
- Single test duration took at least 120 seconds
- All positions were populated via the API in real-time with a latency less than 30 ms
- The number of calculated positions was 98%+
- Average CPU load was below 40%
Performance test results:
WebSockets disabled – only UDP API stream
|RTLS Studio 2.2||RTLS Studio 2.3|
|Maximum Number of Tags for UDP API||300||1200||4x more tracked objects|
WebSockets enabled – REST, WebSockets and UDP API streams
|RTLS Studio 2.2||RTLS Studio 2.3|
|Maximum Number of Tags for UDP API||100||1000||10x more tracked objects|
|Maximum Number of Tags for WebSockets API||100||350||3.5x more tracked objects|
With the newly added support for multicore processors, RTLS Studio 2.3 can now calculate from 3.5 to 10x more tags on the same machine (depending on the API output). Sewio customers can seamlessly expand their RTLS projects and cover new use cases without having to deal with server performance issues.
2. TCP Sync gen2 Release: Leveraging Machine Learning Genetic Algorithm for Improved Flexibility
Synchronization of anchors is an essential part of TDoA location-powered RTLS to ensure reliable and sub-meter position accuracy. RTLS Studio now supports not only UDP Sync but also the TCP Sync gen2 scheme to empower greater flexibility of network deployments as the synchronization can now pass through multiple routers.
The benefits of TCP Sync gen2 include:
- RTLS Studio can be deployed on a different network than the anchors. TCP synchronization passes through the routers
- Auto anchor alignment in TDMA is driven from RTLS Studio entirely independently of the network infrastructure with microsecond precision. No additional precise synchronization hardware, such as IEEE 1588, is needed
- SYNC timeslot per master anchor is reduced by half, which allows you to use more master anchors
- Self-healing and automatic joining of multiple time domains
Anchor Assignment in TDMA Explained
Let’s explain assigning anchors into time slots, known as Time Division Multiple Access (TDMA), in more detail using the RTLS layer model.
TCP SYNC lies within the second layer – see the picture below. To provide positioning coverage among all the anchors, selected anchors (master anchor) must synchronize with the rest via UWB transmission (layer 3). The anchor contains a single UWB radio, so it can either transmit the synchronization or receive blinks from tags at that time. If there are multiple master anchors within the same radio range, they should not transfer at the same time. Otherwise, a radio collision would occur. On the other hand, if they are not at the same range, their transmission could overlap, thereby utilizing over-the-airtime optimally. TCP Sync implements aligning anchors within the TDMA, which is done via a persistent unicast TCP connection with each anchor. Since optimal alignment can become a very complex optimization problem, especially as the number of anchors grows, we have employed an advanced machine-learning-based genetic algorithm to calculate the optimum solution in almost real-time.
Synchronization Layer Model
TCP Sync gen2 also got a new UX interface.
3. RTLS Studio Docker Container: An Operating System-agnostic Solution
Now, RTLS Studio supports deployment via a read-made Docker container, making the RTLS Studio platform OS agnostic. Since the porting of RTLS Studio into different operating systems is not an easy task, due to specific restrictions and dependencies, the Docker is the perfect solution for running it in different environments. As long as your platform can run Docker, you can run RTLS Studio as well. When running the Docker container on a Linux machine, the performance of RTLS Studio is significantly better than when using a virtual machine. For Windows, the performance is comparable with virtual machines.
Let’s sum up Docker’s advantages:
- Wide industrial acceptance (from automotive to aerospace)
- Easy and fast deployment
- Better computational performance than virtualization
- Native integration with the Sewio CI/CD pipeline, providing faster software delivery cycles
The Docker container is available for selected early adopters only.
4. Sensmap Improvements: Better Usability
Apart from the several small improvements, we have brought two major features to Sensmap, our real-time visualization tool, that enhance the user experience.
Export and Import of Anchor Restrictions
Anchor restrictions give additional information to the positioning algorithm. The possibility to export or import anchor restrictions presents a popular option due to the rapid saving of time once you deploy dozens of anchors. It is a powerful tool, especially when used together with RTLS Player.
Here is a use-case. Let’s imagine that the anchors were deployed within the venue and now you need to set masters.
Master Anchors Now Have a Dedicated Marker
Your voice has been heard! It may seem just a detail, but showing a master anchor with a specific marker helps you know its deployment far more quickly.
5. Sectional Initialization: Streaming Large Deployments
If you decide to extend an already large project with 100+ anchors, the initialization of the whole deployment, which would cover not only the newly added anchors but also the original project, might result in a time-consuming task. That is why we have implemented sectional initialization, where the deployment is done consecutively, together with auto anchor deployment, to cut the time of initialization. Anchors can be selected using various filters, e.g., by anchor alias, so you could initialize only anchors within a given room, as shown in the picture.
6. Anchor Firmware Updating from RTLS Studio: Faster and Safer Upgrades
Until now, anchor firmware could be updated only using an external utility. The anchor firmware update feature is now native to RTLS Studio and it includes validators to prevent incorrect updates, making the update process faster, flawless and safer.