IOT CONNECTED
EDGE
DEVICES
LoRaWAN Class A/B/C with adaptive Spreading Factor; LTE-M Cat-M1 and NB-IoT Cat-NB1 with eDRX and PSM; BLE 5.x, Zigbee, Thread, Matter, WirelessHART; sub-µA sleep firmware for multi-year battery life; TFLite Micro edge AI; MQTT, OPC-UA, and SUIT OTA — designed for the deployment environment, not the bench.
The Engineering Problem
IoT Devices Are the Most Deceptively Difficult Class of Embedded Electronics to Get Right at Scale
Power budget, wireless protocol, antenna design, edge processing, enclosure qualification, and platform data model are co-dependent decisions that must all be correct from the first specification. A LoRaWAN node without a power budget accounting for retransmission at actual RSSI won't hit battery life claims. A BLE gateway without antenna isolation for simultaneous BLE/Wi-Fi operation will produce throughput degradation that takes months to trace back to PCB layout.
Matter adoption demands Thread-capable hardware with CSA DCL attestation for multi-ecosystem interoperability. Industrial IoT consolidation drives OPC-UA and standardized data model requirements that legacy hardware can't meet. Edge AI democratization via TFLite Micro and Edge Impulse creates demand for hardware with the MCU performance and NPU acceleration to run meaningful on-device inference.
IoT platform companies, industrial automation vendors, smart building operators, and hardware startups all need an engineering partner who designs from the deployment environment and platform integration requirement outward — and validates end-to-end before the first device ships, not after.
IoT Device & Edge Hardware Categories We Build
From LoRaWAN agricultural sensor nodes to industrial OPC-UA edge gateways to Matter-compliant smart home devices — custom IoT electronics across every protocol, environment, and platform.
Industrial IoT & Edge Computing Hardware
Smart Building & Facilities IoT
Agricultural & Outdoor IoT Devices
Smart Home & Matter IoT Hardware
Healthcare & Clinical IoT Devices
Energy & Utilities IoT Hardware
Fleet, Transport & Logistics IoT
Environmental & Remote Monitoring IoT
Wireless Protocol Selection & Implementation
LoRaWAN Class A/B/C, LTE-M with eDRX/PSM, NB-IoT, BLE 5.x, Zigbee, Thread, Matter, WirelessHART, and Iridium SBD/Swarm satellite — protocol selected based on coverage, power, latency, and platform requirements, not cost alone.
Power Budget & Energy Harvesting
Battery lifetime calculated from transmit duty cycle, MCU sleep current, sensor measurement energy, and deployment temperature before hardware is selected. Li-SOCl2 for extended temperature, solar MPPT for outdoor perpetual operation, sub-µA sleep with hardware interrupt wake, and supercapacitor buffering for high-pulse transmit.
Edge AI & Local Intelligence
TFLite Micro on Cortex-M4F/M7 for on-device anomaly detection, Edge Impulse training pipeline integration, ONNX Runtime on Cortex-A edge hardware, threshold-based selective transmission firmware, and autonomous local PID control for operation without cloud connectivity.
Platform & Cloud Integration
MQTT 3.1.1/5.0 with TLS 1.3 for AWS IoT Core, Azure IoT Hub, and Google Cloud IoT; OPC-UA server with SCADA information model; OSIsoft PI Connector; Kafka producer firmware; InfluxDB line protocol; and device shadow/twin synchronization.
IoT Security Architecture
ATECC608B/SE050 hardware security elements for device identity and ECDH key exchange, X.509 manufacturing provisioning, TLS 1.3 mutual authentication, secure boot with hardware root of trust, PSA Certified Level 1/2 architecture, and IETF SUIT RFC 9019 signed OTA with rollback.
Antenna & RF Engineering
Chip, PCB trace, and external antenna selection for form factor, enclosure material, and frequency band; co-existence design for simultaneous BLE/Wi-Fi or LoRaWAN/NB-IoT; SAR compliance for worn devices; TRP/TIS radiated measurement on representative enclosures; and antenna matching network design.
Industrial Protocol Bridging
Modbus RTU/TCP to MQTT/OPC-UA bridging with RS-485 bus sharing for SCADA coexistence, HART 7 to cloud gateway, EtherNet/IP and PROFINET adapter firmware, BACnet to MQTT, DNP3 to cloud, J1939 to cellular IoT, and SDI-12 to LoRaWAN for agricultural networks.
OTA & Fleet Management
Delta OTA to minimize transfer size for LoRaWAN and satellite links, cryptographically signed firmware verification, staged rollout with automatic rollback on boot failure, zero-touch provisioning, and battery state-of-health monitoring across deployed fleets.
Chipsets & Platforms
Platforms, Protocols & Standards
Tested silicon and proven stacks — no experimental platform dependencies.
Industrial IoT, Edge Computing & Protocol Bridging
Modbus to OPC-UA. HART to Cloud. ATEX Zone 2. RS-485 Bus Sharing.
Industrial IoT must coexist with legacy Modbus/HART/EtherNet/IP infrastructure, operate in ATEX Zone 2 areas, deliver to OSIsoft PI historians and Ignition SCADA, and never disrupt existing automation communications. A Modbus-to-MQTT bridge must manage polling schedule collisions on shared RS-485 buses, prevent TCP connection resets, minimize cellular payload size, and buffer data through uplink gaps — all simultaneously.
Ultra-Low-Power, Long-Life & Remote Deployment IoT
LoRaWAN Spreading Factor Analysis. Li-SOCl2 at −15°C. Solar at 52°N.
Remote, battery-dependent IoT requires a power budget accounting for every µA of sleep current, every mJ of sensor energy, every retransmission at actual installation RSSI, and every degree of battery chemistry temperature derating. LoRaWAN SF selection is a quantitative trade-off — air time and energy increase ~4× per SF step while link budget improves ~2.5 dB — that must be solved for the actual installation distance and network conditions.
Matter, Thread & Smart Home IoT Hardware
CSA DCL Attestation. Multi-Ecosystem Interoperability. Thread Border Router Integration.
Matter certification requires qualified chip platforms (EFR32MG24, nRF5340, ESP32), Thread border router integration, QR code/NFC commissioning, and CSA DCL device attestation. Stable interoperability across Apple HomeKit, Google Home, Amazon Alexa, and Samsung SmartThings in real-world Thread mesh conditions — not just certification test environments — requires experience with the specific edge cases certification testing doesn't cover.
Why IoT Engineering Teams Choose Ankh
Power Budget Engineering Before Hardware Selection
Battery life is determined by MCU sleep current, radio attach and transmit current, sensor measurement energy, and deployment temperature — not tuned in firmware after hardware is built. We build the power budget at the first architecture review and validate it against the actual network and environment, not a bench at room temperature with full signal.
Wireless Protocol Selection as a System Engineering Decision
LoRaWAN, LTE-M, NB-IoT, BLE, Thread, and satellite are not equivalent options ranked by cost. A LoRaWAN Class A device requiring frequent downlink commands will have unacceptable latency. An NB-IoT device deployed where the carrier hasn't deployed NB-IoT won't connect. We select protocols based on coverage, power, latency, and platform analysis — before a schematic is drawn.
Platform Integration Built into the Device Architecture
A device uploading data in the wrong format to the wrong cloud API is a pilot that never scales. We design the MQTT topic structure, OPC-UA information model, and OTA payload format to match the target platform — and validate end-to-end before the first device ships, not in a post-delivery middleware project.
IoT Security Designed In, Not Bolted On
Hardware security element provisioned at manufacturing, TLS mutual authentication, signed OTA verification, and secure boot from hardware root of trust — all first-order architecture constraints. IoT security retrofitted after the connectivity architecture is built requires a fundamental redesign we design to avoid.
Industrial Edge Gateway — 340-Facility Manufacturing IoT Platform
A manufacturing IoT platform needed a DIN-rail edge gateway bridging legacy Modbus RTU, EtherNet/IP, and PROFINET to cloud across 340 facilities in 14 countries — IECEx Zone 2 Ex ec certified, without disrupting existing SCADA polling on shared RS-485 buses.
Cortex-A53 quad-core edge compute with 4-port RS-485 Modbus RTU for 128 instruments, EtherNet/IP and PROFINET adapter firmware, OPC-UA aggregation server with ISA-95 information model, LTE Cat-1 with eSIM failover, IECEx Zone 2 Ex ec, −40°C to +70°C, AWS IoT Greengrass and OSIsoft PI integration.
- 340 facilities deployed across 14 countries in 9-month programme
- 99.94% fleet uptime in first year of operation
- IECEx Zone 2 Ex ec certification achieved
- Zero SCADA polling disruption on shared Modbus RS-485 networks
LoRaWAN Soil Monitoring Node — 6.2-Year Battery Life, 2,400 Units Deployed
A precision agriculture company needed a LoRaWAN soil monitoring node with 6+ year battery life down to −15°C, validated solar-assisted operation at 52°N winter, and REST API integration with their irrigation scheduling platform.
Solar-assisted LoRaWAN Class A with TDR soil moisture at 10/30/60cm depths, adaptive SF selection, 1W solar MPPT validated at 52°N winter, sub-2µA sleep, IP67. 6.2-year battery backup life measured at −15°C in thermal chamber.
- 6.2-year battery backup life validated at −15°C in thermal chamber
- 99.1% data capture rate across 2,400 nodes, first growing season
- Solar charge balance maintained through 4-week overcast at 52°N
- ±1.1% VWC soil moisture accuracy across 3 soil types
Building an IoT connected device or edge hardware platform? Let's connect it.
Power budget engineering from deployment environment outward. Wireless protocol selection as a system engineering decision. Full-stack delivery from hardware through firmware through platform integration through OTA fleet management.
Quote Your IoT Project