01 Background: how industrial wireless sensor networks evolved
An industrial wireless sensor network is formed by many sensing nodes with acquisition, compute and wireless capability organising themselves into a network to perform monitoring and control. It suits complex industrial environments, offering strong interference resistance, very low energy use and real-time communication, while keeping deployment and operation flexible and inexpensive.
The first generation of industrial wireless protocols
The main early technologies were WirelessHART, ISA100.11a and WIA-PA. That generation proved wireless was viable for process value acquisition in process industries, but shared common problems: proprietary protocols, closed ecosystems and difficulty integrating with IT networks.
Why IEEE 802.15.4e TSCH matters
The TSCH MAC mode introduced in the IEEE 802.15.4e specification won broad support from both industry and academia. In an IEEE 802.15.4e TSCH network:
- Nodes coordinate their state through precise time synchronisation, cutting idle listening time and therefore power consumption
- Channel hopping between nodes improves the reliability of wireless communication
- Scheduling is planned centrally by the network manager, removing the uncertainty of random access
Research and practice show that IEEE 802.15.4e TSCH can bring industrial wireless communication close to the 99.999% reliability level of wired communication.
The arrival of the 6TiSCH standard
The IETF developed and, in 2021, completed 6TiSCH — a productisable protocol and architecture standard for deterministic industrial wireless mesh networks. It merges IPv6 with IEEE 802.15.4e TSCH so that industrial IoT nodes connect seamlessly into industrial internet and industrial IoT systems.
Choosing 6TiSCH as its foundation means AIMesh stood on an international standard from the outset, rather than on a closed proprietary protocol.
02 Core technology
AISENZ has worked continuously on next-generation wireless sensor network technology, producing AIMesh — its own industrial wireless sensor network protocol — together with the associated industrial IoT products. AIMesh optimises the 6TiSCH stack and architecture, introduces new physical-layer modulation, designs a distinctive distributed intelligent scheduling algorithm and strengthens network management and security, building a commercially deployable real-time deterministic industrial wireless sensor network.
Physical layer
FLRC/FEC modulation, with a single-hop distance 2× that of DSSS or GFSK.
MAC layer
Follows the IEEE 802.15.4e TSCH standard with up to 45 hopping channels, effectively reducing message collisions.
Logical link layer
Adapts upward to IPv6 and downward to IEEE 802.15.4e, with support for data compression and packet fragmentation and reassembly.
Network layer
Compatible with the IPv6 layer, carrying kilobyte-scale IP packets, with RPL as the routing protocol.
Scheduling algorithm
A distinctive resource allocation and scheduling algorithm supporting real-time, reliable uplink and downlink transmission of several data types: high-frequency small packets, low-frequency large packets, burst packets, alarm packets and control commands.
Security and frequency bands
Link-layer security follows the IEEE 802.15.4e mechanism; IP-layer security uses DTLS. The protocol runs in either the SubG or 2.4G band, meeting licensing requirements across regions worldwide.
Scale and operations
Network scale expands and contracts freely, with multi-access-point deployment for elastic growth. Web-based visual remote management covers topology, configuration, performance, alarm display and operations.
03 Measured indicators
The following were collected on the standard test configuration: 100 nodes, 3-hop topology, one packet per 30 seconds, multi-storey deployment.
- Whole-network formation or recovery: under 3 minutes
- Single node join time: under 30 seconds
- End-to-end reliability: above 99.99%
- End-to-end latency: under 1 second
- Battery life, low-power non-routing nodes: 5–10 years, by reporting frequency
These are not laboratory peaks. They are engineering test values taken with a 3-hop topology across multiple storeys, which covers most real industrial sites.
04 Network architecture
An AIMesh network is built from three kinds of node, each with its own role:
Mote nodes
Mote nodes are the industrial IoT wireless nodes. By role they can be routing or non-routing, low-power or mains-powered. Low-power non-routing nodes focus purely on data acquisition and run five to ten years on battery; routing nodes carry relay duty and are normally powered from mains or a long-life supply.
AP nodes
The AP node bridges the wireless and wired networks. Each AP forms a wireless subnet together with a number of mote nodes. Deploying multiple APs extends the network and improves signal coverage, so AIMesh scales from a single floor to a large plant.
Gateway nodes
The gateway node is the operational and management anchor for the whole network, handling protocol conversion, data modelling and edge computing, and exposing a unified data access service. Northbound it connects to SCADA, MES or an IoT platform; southbound it manages several AP subnets.
Network capacity
AIMesh uses multiple AP nodes to extend the network and improve signal coverage. One AP subnet accommodates 100 mote nodes, and a network scales linearly to 255 subnets, enough to carry dense monitoring across a large refinery complex, campus or data centre.
05 Typical applications
AIMesh has been validated across a range of industrial settings. Four representative applications:
Oil and gas production IoT
AIMesh replaces WirelessHART for production condition monitoring across oil and gas exploration, extraction, storage, transport and refining, covering temperature and pressure variation, flow, level and energy consumption. It supports second-level data refresh with reliability above 99.99% and five to ten years of battery life on low-power instruments.
Condition monitoring for process industry equipment
Rotating machines, reciprocating equipment and pumps in process industries need online condition monitoring and diagnosis. Their condition data covers vibration, temperature, pressure, eddy current, acceleration and photoelectric sensing. AIMesh supports acquisition from the relevant low-power sensors and real-time self-organising transmission with reliability above 99.99%, giving predictive maintenance models high-quality continuous data.
Cloud data centre monitoring
In data centres, AIMesh focuses on environmental monitoring, infrastructure management and energy efficiency. In electrically noisy environments such as dense server halls it keeps packet loss below one in ten thousand. It supports remote monitoring of hundreds of UPS units, cooling systems and power distribution units, feeding dynamic cooling strategy and load balancing and reducing PUE directly.
PV and wind farm monitoring
AIMesh wireless nodes acquire voltage, current and temperature from generation equipment in real time and forward it to the management platform for remote monitoring and fault warning. Second-level data refresh, reliability above 99.99% and fast self-organising recovery make it particularly suited to renewable sites with scattered points and difficult terrain.
One AIMesh protocol spans four entirely different kinds of site — oil and gas, process industry, data centres and renewables. That is the fundamental advantage of a standardised protocol over a purpose-built one.Frequently asked questions
How does AIMesh relate to earlier protocols such as WirelessHART, ISA100.11a and WIA-PA?
That generation proved wireless was viable for process value acquisition in process industries, but shared proprietary protocols, closed ecosystems and difficulty integrating with IT networks. AIMesh is built on 6TiSCH, which the IETF completed in 2021, merging IPv6 with IEEE 802.15.4e TSCH so industrial IoT nodes connect seamlessly into the industrial internet — standing on an international standard from the outset rather than a closed proprietary protocol.
Under what conditions were the AIMesh indicators measured?
On the standard test configuration: 100 nodes, 3-hop topology, one packet per 30 seconds, multi-storey deployment. The figures are whole-network formation or recovery under 3 minutes, single node join under 30 seconds, end-to-end reliability above 99.99%, end-to-end latency under 1 second, and five to ten years of battery life on low-power non-routing nodes. These are engineering test values across a 3-hop multi-storey topology, not laboratory peaks.
What kinds of node make up an AIMesh network?
Three. Mote nodes are the wireless sensing nodes — routing or non-routing, low-power or mains-powered; low-power non-routing nodes focus on acquisition and run five to ten years on battery, while routing nodes carry relay duty on mains or a long-life supply. AP nodes bridge wireless and wired, each forming a subnet with a number of motes, with multiple APs extending reach and coverage. Gateway nodes are the operational and management anchor, handling protocol conversion, data modelling and edge computing, connecting northbound to SCADA, MES or an IoT platform and managing several AP subnets southbound.
How many nodes can one network hold?
One AP subnet holds 100 mote nodes and a network scales linearly to 255 subnets, reached by deploying multiple AP nodes to extend the network and improve coverage — enough for dense monitoring across a large refinery complex, campus or data centre.
Which frequency bands does AIMesh run in?
Either SubG or 2.4G, so it meets licensing requirements across regions worldwide.
Standards & references
The specifications and authoritative entries behind the protocols and standards this article discusses.