- Manufacturers
- Utilities
- Facilities
- Agriculture and field operations
- Product builders
Catalogue › Hardware & Equipment Marketplace › IoT, Edge Computing & Embedded Systems
Hardware & Equipment Marketplace
IoT, Edge Computing & Embedded Systems
Sensors, gateways, industrial computers and embedded components for connected operations, monitoring, automation and edge intelligence.
- Real-time operational visibility
- Local processing and resilience
- Reduced manual inspection
- Connected equipment and data services
- Sensors and industrial gateways
- Edge computers and AI accelerators
- Cellular, LoRaWAN, Wi-Fi and wired connectivity
- Microcontrollers, SBCs and embedded modules
- Device management and secure provisioning
- Enclosures, power, integration and deployment
- Approved bill of materials or manufacturer part selection
- Technical datasheets and compatibility record
- Warranty, authenticity and regulatory documentation
- Configuration, staging or asset tagging when purchased
- Shipping, installation and acceptance records
- Support, replacement and end-of-life information
- Monitoring starter kit
- Industrial edge gateway
- Remote asset kit
- Environmental sensing
- Edge AI appliance
- Custom embedded design
- Brand, model and technical specification
- Quantity, stock and supplier price validity
- Configuration, licenses and accessories
- Warranty and support level
- Freight, installation, taxes and duties
- Compatibility and lifecycle availability
Content on this page comes from the governed ARRIX catalogue record HW-09; pricing is confirmed only through a reviewed quotation.
What This Is
IoT, Edge Computing & Embedded Systems
Connected device and edge systems are the sensors, gateways, controllers and small ruggedised computers that measure what is happening in the physical world and process it close to where it occurs.
This is technology placed where the work actually happens - in a vehicle, on a machine, in a cold store, on a wall, in a field. Sensors measure something: temperature, movement, level, vibration, position, energy. A gateway collects those readings and passes them on. An edge computer does the thinking locally instead of sending everything to a distant server, which matters when the connection is poor, when the response must be immediate, or when the volume of data would be uneconomic to transmit. The pattern is always the same - measure, decide, act - and the value lies in knowing something without a person having to go and look.
Problems are seen before they become failures
Continuous measurement with thresholds and alerting surfaces drift and abnormality while there is still time to act.
Fewer wasted journeys and inspections
Remote monitoring replaces routine physical checks, and visits are made when the data says they are needed.
Local decisions survive a lost connection
Edge processing keeps control and safety logic running on site when the network or internet is unavailable.
Lower data transmission cost
Filtering and analysing locally means conclusions and exceptions are sent rather than raw streams.
Evidence of conditions over time
Continuous logging produces the records that food safety, pharmaceutical, environmental and insurance obligations increasingly expect.
Maintenance planned rather than reactive
Vibration, temperature and runtime data allow servicing to be scheduled on condition instead of on the calendar or on breakdown.
Why It Matters
The problems this answers
Connected device projects fail on operations rather than on technology. The sensors work; the difficulty is powering, connecting, mounting, securing and maintaining hundreds of small devices in places that are inconvenient to reach, and then making someone accountable for the readings they produce. The economics turn on three questions: what decision will change as a result of the data, how the devices are powered and connected, and who maintains them once the project team has moved on. Edge computing has grown because sending everything to a central system is often impractical - connections at the edge are intermittent, some responses cannot wait for a round trip, and continuous video or high-frequency sensor data is expensive to move. Processing locally and sending conclusions rather than raw data addresses all three. This is the least commoditised of the hardware families: deployments are designed rather than selected, and a pilot before a rollout is the norm rather than a precaution.
- Conditions that matter - temperature, level, presence, vibration, energy - are known only when someone checks in person
- Equipment failing without warning, causing unplanned downtime and emergency callouts
- Data being generated at sites with poor, expensive or intermittent connectivity
- Responses that must happen in milliseconds and cannot wait for a round trip to a central system
- Manual readings, paper records and inspection rounds consuming staff time and still arriving too late
- Compliance requiring continuous evidence of conditions rather than periodic spot checks
| Organisation | Need | What the equipment does | Outcome |
|---|---|---|---|
| A cold storage or food business | Continuous proof that temperature stayed within range | Wireless temperature sensors with a gateway and alerting | Excursions raise an alarm in time to save stock, and records satisfy an auditor. |
| A manufacturer | Warning before a machine fails | Vibration and temperature sensors with edge analysis on the line | Servicing is scheduled on condition and unplanned stoppages fall. |
| A facilities team | Understanding where energy is actually consumed | Sub-metering with a gateway and reporting | Consumption is attributed to areas and equipment, so reduction is targeted rather than guessed. |
| A utility or agricultural operator | Readings from remote sites with no network | Low-power wide-area sensors or cellular gateways with local buffering | Distant assets report without a visit, and data survives connection gaps. |
| A retail or logistics operation | Analysing camera or sensor data without moving it all off site | Edge computer performing analysis locally | Only results are transmitted, cutting bandwidth cost and keeping raw data on site. |
| An equipment manufacturer | Supporting installed products remotely | Embedded controller with secure connectivity and remote update | Faults are diagnosed without a visit and firmware is maintained across the installed base. |
How To Choose
Types, and when each one is the right answer
Sizing matters more than brand here. ARRIX specifies to the requirement rather than to a fixed model list.
Environmental sensors
Temperature, humidity, air quality, water or light must be measured continuously - the most common and lowest-risk entry point.
Condition monitoring sensors
Vibration, current, acoustic or thermal signatures indicate the health of rotating or powered equipment.
Presence, level and flow sensors
Occupancy, tank contents, stock level or throughput need to be known without inspection.
Asset tracking devices
The location or movement of equipment, vehicles or containers must be known, indoors or across a region.
Metering and energy sensors
Consumption must be attributed to specific circuits, areas or machines before it can be managed.
Short-range wireless gateway
Many nearby low-power sensors need to be collected and forwarded from a single point.
Low-power wide-area gateway
Sensors are spread over a large area, send small readings infrequently and must run for years on a battery.
Cellular gateway or router
The site has no fixed connection, or a resilient independent path is needed.
Industrial edge computer
Local processing must run in temperature, vibration, dust or electrically noisy conditions unsuitable for office equipment.
Edge server with acceleration
Video analysis or machine learning inference must run on site rather than in a central location.
Programmable logic controller and industrial control
Deterministic machine control is required, with reliability and safety expectations that general computing does not meet.
Single-board and embedded computers
Prototyping, low-volume products or embedding computing inside another piece of equipment.
Rugged human-machine interface panels
Operators need a durable local interface on a factory floor, vehicle or outdoor installation.
Protocol gateways and converters
Existing industrial equipment speaks an older protocol that must be bridged to a modern network.
What decides the choice
- The decision the data will change - a deployment that produces readings nobody acts on is an expense rather than an investment
- Power at each device: mains, power over the data cabling, solar or battery, and if battery, the expected life and the cost of replacing it at scale
- Connectivity available at the location and its reliability, which usually decides the technology before anything else does
- Environmental conditions - temperature range, moisture, dust, vibration, corrosion, electrical noise
- Required response time, which determines what must be processed locally rather than centrally
- Data volume and transmission cost, particularly on cellular or metered connections
- Security and update path over a life that is frequently ten years or more
- Manageability at scale, since a hundred devices cannot be maintained the way one can
- Standards and interoperability, to avoid a deployment that only one supplier can ever extend
Specification Detail
For whoever has to sign it off
Open only what you need. Nothing here is hidden from print or from a browser without JavaScript.
Specifications that actually decide the outcome
| Specification | Weight | What it means | Why it matters | Specify higher when |
|---|---|---|---|---|
| Operating temperature range | Critical | The ambient range within which the device functions reliably. | Office-rated equipment fails in cold stores, plant rooms, vehicles and outdoor cabinets. | Deployment is outdoors, industrial, refrigerated or unconditioned. |
| Ingress protection rating | Critical | Sealing against dust and water. | Determines survival in wash-down, outdoor and dusty environments. | Exposed to weather, cleaning or particulates. |
| Power source and consumption | Critical | How the device is powered and how much it draws. | Battery life at scale is an operational cost; replacing hundreds of cells is a recurring project. | Devices are numerous, remote or difficult to reach. |
| Connectivity technology and range | Critical | The radio or wired method and the distance it covers. | Range, penetration, power draw and data capacity trade against each other; no option is best at all four. | Devices are distant, underground, inside metal, or must run for years on a battery. |
| Measurement accuracy and resolution | Critical | How closely and how finely the sensor reflects reality. | Compliance and control decisions depend on accuracy that can be evidenced and, where required, calibrated. | Readings support a regulatory record or a control action. |
| Sampling and reporting interval | Important | How often the device measures and transmits. | Drives battery life and data volume, and determines how quickly a problem is noticed. | Conditions change quickly or an excursion must be caught early. |
| Local processing capability | Important | Compute, memory and any acceleration available on the device. | Decides what analysis can happen on site rather than centrally. | Video, inference or fast local control is involved. |
| Local storage and buffering | Important | Data retained when the connection is unavailable. | Without buffering, a connection outage becomes a permanent gap in the record. | Connectivity is intermittent or the record must be complete. |
| Security features | Critical | Secure boot, encrypted communication, device identity and authenticated updates. | These devices are numerous, long-lived, physically exposed and frequently forgotten - which is precisely the profile attackers look for. | Devices are internet-connected, physically accessible or control something. |
| Firmware update mechanism | Critical | How software is updated across the estate once deployed. | A device that cannot be updated remotely becomes a permanent liability at scale. | Devices are numerous, remote or expected to last many years. |
| Mounting and physical form | Important | Rail mount, panel, enclosure, magnetic or fixed installation. | Installation practicality frequently decides feasibility more than any electrical specification. | Space is constrained or the mounting surface is awkward. |
| Certification and approvals | Specialized | Radio, electrical, hazardous-area and industry-specific approvals. | Some environments legally require certified equipment, and non-compliant devices cannot be installed. | Deployment is in hazardous, medical, marine or regulated settings. |
What it has to work with
- Sensors, gateways and platforms must speak a common protocol; standards give basic interoperability but advanced features are often ecosystem-specific
- Low-power wide-area technologies depend on network availability - public coverage, a private gateway, or neither - which must be confirmed at the actual location
- Industrial equipment frequently uses older serial or fieldbus protocols and requires a converter or gateway to reach a modern network
- Cellular deployments depend on coverage, on the operator, and on the generation of network supported - older network generations are being withdrawn on operator timetables
- Edge computers must run the software stack intended for them, and processor architecture and accelerator support determine what will actually run
- Devices should be segmented from general computing, which requires the network to be capable of that separation
What it costs to own, not just to buy
- Hardware is often the smallest element; installation, mounting, commissioning and integration dominate the initial cost
- Connectivity subscriptions recur per device and become significant at scale
- Platform, storage and analytics charges are usually per device or per data volume and grow with the deployment
- Battery replacement across a large estate is a recurring operational project, not an incidental cost
- Calibration is required where readings support compliance, and it recurs
- Long-lived devices require a maintained update path for their full service life, which is an ongoing commitment rather than a one-off
- The most commonly omitted cost is the human one: someone must monitor, interpret and act on what the system reports
Risks and things worth knowing first
- Deployments that produce data nobody acts upon are the most common failure in this family - the decision must precede the sensor
- Connected devices are a recognised weak point in network security: numerous, physically exposed, long-lived and easily forgotten
- Devices that cannot be updated become permanent liabilities and are difficult to justify replacing individually
- Battery replacement at scale is frequently underestimated at design and painful in service
- Vendor lock-in is a real risk where proprietary platforms hold the data and the device management
- Connectivity assumptions made at the office are often wrong at the installation point; coverage must be tested where the device will actually sit
- Regulation - radio approvals, hazardous areas, data protection where people are observed - constrains deployments and must be established early
- Pilots succeed at ten devices and reveal their real costs at a thousand
Mistakes that are common and expensive
- Starting from the technology instead of from the decision the data should change
- Assuming connectivity exists at the installation point without testing it there
- Overlooking battery replacement cost across the full estate
- Using office-rated equipment in industrial, refrigerated or outdoor conditions
- Deploying devices with default credentials, no segmentation and no update path
- Sampling far more frequently than any decision requires, and paying for the data and the batteries
- Choosing a closed platform and discovering the data cannot be taken elsewhere
- Scaling straight from a successful pilot without re-examining the per-device operational cost
When you may not need this at all
If the condition being measured is already visible to someone who is there anyway, if nothing would be done differently when the reading changed, or if the equipment concerned is inexpensive to replace and its failure is not disruptive, instrumentation is unlikely to repay its cost. Small deployments can also be served by simple standalone loggers rather than a connected platform. The case strengthens where sites are hard to reach, where failure is expensive, where evidence must be continuous, or where the response must be faster than a person can provide.
Common Questions
Answered plainly
Where should an IoT project start?
With the decision, not the device. The only deployments that repay their cost are those where somebody will do something different because of the reading. Name the decision, name the person who acts on it, and the sensor specification, sampling rate and connectivity choice all follow from that. Projects that begin with the hardware tend to produce dashboards nobody opens.
What IoT devices do you stock?
ARRIX does not hold a fixed device list at family level - this is the least commoditised of the hardware families, and deployments are designed rather than picked from a shelf. What is being measured, where, in what conditions, with what power and what connectivity determines the specification. Describe the situation and ARRIX will design to it and quote.
What is edge computing, in practical terms?
Processing data where it is produced instead of sending it all somewhere else first. It is used for three practical reasons: the connection at the site is poor or expensive, the response has to be faster than a round trip allows, or the volume of data - video especially - would cost more to transmit than the answer is worth. The device sends conclusions rather than raw streams.
How long do sensor batteries last?
Anywhere from months to several years, and the figure is decided mostly by how often the device transmits and which radio technology it uses. A low-power sensor reporting every fifteen minutes behaves very differently from one reporting every ten seconds. The important question is not the life of one battery but the cost of replacing them across the whole estate, which is a recurring operational project once the count runs into the hundreds.
Are IoT devices a security risk?
They have the profile attackers prefer: numerous, physically accessible, long-lived, and frequently forgotten after installation. That is manageable, but only deliberately - segmented from general computing, credentials changed at commissioning, and a documented remote update path for the full service life. A device that cannot be updated should be treated as a liability from the day it is installed.
Do I need cellular, Wi-Fi or a low-power network?
It depends on distance, data volume and power. Wi-Fi suits mains-powered devices within an existing network. Cellular suits sites with no fixed connection or where an independent path is wanted. Low-power wide-area technologies suit small, infrequent readings from battery devices spread over a large area. The one thing that cannot be assumed is coverage - it has to be tested at the exact spots where devices will sit.
Our pilot worked. Why would a full rollout be different?
Because most of the cost in this family is per-device operational rather than per-device capital. Ten devices can be installed by hand, watched informally and have their batteries changed in an afternoon. A thousand cannot. Before scaling, the figures worth re-examining are installation labour, connectivity subscriptions, platform charges per device, battery and calibration cycles, and who is accountable for the alerts.
Request A Quotation
Tell ARRIX the situation, not the part number
ARRIX sources this family to requirement. A specified quotation is faster than a catalogue search, and these are the questions it answers.
- What are you trying to measure or control, and what decision would change as a result?
- How many measurement points, and how far apart are they?
- What is the environment - indoor, outdoor, refrigerated, industrial, hazardous, vehicle?
- What connectivity exists at those exact locations - wired, wireless, cellular or none?
- Is mains power available at each point, or must devices run on battery or solar?
- How quickly must you know when something changes - seconds, minutes or hours?
- Does anything need to keep working when the connection is lost?
- Is there existing equipment or machinery this must read from, and what protocol does it use?
- Is there a compliance or audit requirement for the records?
- Who will receive the alerts and act on them?
- Is this a pilot or a full deployment, and what is the eventual scale?