How to Choose a LoRaWAN Temperature & Humidity Sensor (2026)

How to Choose a LoRaWAN Temperature and Humidity Sensor: A 2026 Buyer’s Checklist for India

Short Answer

To choose a LoRaWAN temperature and humidity sensor, start from where it will be mounted, not from the datasheet. Check these eight things, in this order:

1. Environment. A room, cold room, freezer or outdoor site decides the form factor.
2. Accuracy across your range. Read the accuracy at your operating temperature, not at 25 °C.
3. Humidity and condensation. Most sensors are rated non-condensing. Many cold rooms are not.
4. Built-in sensor or external probe. Freezers, ducts and process equipment need a probe.
5. Battery life at your reporting interval. Ignore the headline “10 years” figure.
6. Radio version for India. You need IN865 firmware and WPC type approval for the exact model.
7. Data continuity. Find out what happens to readings when the gateway or internet connection drops.
8. Calibration evidence. You need a traceable certificate if an auditor will ever look at the data.

Then run a two-to-four-week pilot before you buy at scale.

Why Choosing Is Harder Than It Looks

Open five LoRaWAN temperature and humidity sensor datasheets and they look almost identical: ±0.3 °C, ±3 %RH, IP67, “up to 10 years” battery life. The differences that decide whether your deployment works sit in the footnotes:

• the accuracy outside 0–70 °C
• the words “non-condensing”
• the reporting interval behind the battery claim
• whether the unit you are buying is the EU868 or the IN865 version

This guide walks through those footnotes in the order they matter. If you want the basics of how these sensors measure and transmit data first, read How a LoRaWAN Wireless Temperature and Humidity Sensor Works.

1. Match the Sensor to the Environment

The mounting environment decides the sensor type before any specification does. A sensor that is perfect for an office will fail in a blast freezer, and a rugged outdoor unit is overkill on a server rack.

Write down the minimum and maximum temperature at the mounting point, and whether water or condensation can reach it. That single line rules out half the market.

2. Accuracy That Matters Across Your Range

A sensor’s headline accuracy usually applies to only part of its range, so check the figure at your actual operating temperature.

Most datasheets give a tighter accuracy band at room temperature and a looser one below 0 °C or above 90 %RH. Our own MacSync LTS X1 is a typical example. It is specified at:

• Temperature: ±0.3 °C from 0 to +70 °C, and ±0.6 °C from −30 to 0 °C.
• Humidity: ±3 %RH between 10 and 90 %RH, and ±5 %RH outside that band.

In a −20 °C freezer, the relevant figure is ±0.6 °C, not ±0.3 °C.

Three more things to check:

• Chip accuracy is not device accuracy. A sensing IC may be rated ±0.2 °C. Once it sits inside an enclosure next to a battery and radio, the finished device is usually rated looser. Always use the device figure.
• Resolution is not accuracy. A reading of 4.27 °C (0.01 °C resolution) can still be 0.5 °C away from the true value.
• Response time matters for alarms. A sensor in a thick enclosure can take several minutes to register a door-open spike.

Rule-of-thumb accuracy targets:

3. Cold Rooms and Condensation: The Humidity Problem Nobody Specifies

Most temperature and humidity sensors, including ours, are specified for non-condensing conditions, and many cold rooms regularly condense.

Every door opening lets warm, humid air into a cold space. Defrost cycles push moisture through the room. When water condenses on a capacitive humidity element, the reading jumps to around 100 %RH and can stay there until the element dries out. The temperature reading is usually unaffected; the humidity data is what suffers.

How to handle it:

• Mount carefully. Keep the sensor away from the door and away from the evaporator fan’s discharge. Put it where air circulates, with the sensor opening facing down.
• Protect the element. For persistently wet rooms, use a probe with a sintered or PTFE filter rather than an open vent.
• Ask about a heater. Some sensing ICs include a built-in heater that can drive off moisture. Ask whether the firmware actually uses it.
• Read humidity as a trend. In a condensing room, dew point and humidity trend over hours are more useful than any single reading.

If humidity in a condensing room is a compliance parameter rather than an indicator, specify a sensor rated for condensing service. Expect to pay more for it.

4. Built-in Sensor or External Probe?

If the measuring point is colder than about −30 °C, inside metal, or in a process line, use an external probe and keep the sensor body outside.

Putting the whole sensor inside a freezer causes two problems:

1. Battery capacity falls at low temperature, so battery life shrinks.
2. A metal freezer box weakens the radio signal, so packets go missing.

A probe passed through a cable gland or a sealed door port keeps the battery and radio in the warm, open air while the tip measures the cold.

Probe options:

• Digital temperature/humidity probes suit cold rooms and ducts within roughly −40 to +80 °C.
• PT100/PT1000 RTD probes suit deep freezers, process lines and high temperatures. They are the standard choice where accuracy must hold across a wide range.

Do not run a probe cable under a door gasket. The gasket crushes the cable over time and leaks cold air. Use a proper pass-through.

5. Battery Life: Read the Reporting Interval, Not the Headline

Battery life depends mainly on how often the sensor reports, so a “10-year” claim means nothing without its interval and spreading-factor assumptions.

The MacSync LTS X1, with a 19 Ah lithium thionyl chloride (Li-SOCl₂) cell, is rated at:

Three factors shorten that further:

• Weak signal. Poor gateway coverage forces a higher spreading factor (SF10 to SF12), and each packet then costs many times more energy.
• Confirmed uplinks. Asking the network to acknowledge every packet uses extra energy.
• Cold. Low temperatures reduce usable battery capacity.

Practical tip: use threshold-based reporting. The sensor reports every 15 or 30 minutes in normal conditions, but sends an immediate alert when a limit is crossed. You get fast excursion alerts without paying for 5-minute reporting all year.

If mains or DC power is available (control panels, telecom shelters, plant rooms), choose a powered variant. It removes battery replacement from your maintenance plan entirely.

Estimate your own figure with the LoRaWAN Battery Life Calculator.

6. Buying for India: IN865 and WPC

A LoRaWAN sensor deployed in India must run the IN865 channel plan and should hold WPC Equipment Type Approval for the exact model.

India’s licence-exempt band for short-range devices was extended to 865–868 MHz in December 2021 (G.S.R. 853(E)). LoRaWAN networks in India use the IN865 regional channel plan within this band.

Many sensors sold online, including stock imported through European distributors, are EU868 versions. An EU868 unit uses a different channel plan and will not join an IN865 network unless the vendor supplies IN865 firmware for it. Even then, it still needs Indian type approval.

Before you buy, confirm:

• IN865 is listed in the frequency bands for the unit you are buying, not just for the product family.
• The supplier holds WPC Equipment Type Approval for that model.
• There is local support and a clear return (RMA) process.

7. “LoRa” Is Not “LoRaWAN”: Check Compatibility

A sensor sold as “LoRa” may only talk to its maker’s own receiver, while a LoRaWAN sensor works with any standard gateway and network server.

Marketplace listings often use the two terms loosely. Private LoRa systems can work well, but they lock you into one vendor’s receiver and software.

A true LoRaWAN sensor should state:

• LoRaWAN version (1.0.3 or 1.0.4 is typical)
• Class A operation for battery devices
• OTAA activation
• A published payload decoder
• Tested compatibility with network servers such as ChirpStack or The Things Stack

If the datasheet doesn’t mention a LoRaWAN version or a decoder, ask. For the full difference, see LoRa vs LoRaWAN: What’s the Real Difference?

8. What Happens When the Gateway Goes Offline?

By default, a LoRaWAN sensor’s readings are lost if no gateway hears them or the gateway’s internet connection is down. Check the sensor’s local logging and the gateway’s buffering before you buy.

Battery LoRaWAN sensors normally send unconfirmed uplinks: they transmit and go back to sleep. If the gateway is offline during that window, the reading never reaches your dashboard.

For audit-grade records, two safeguards matter:

• Local logging on the sensor, with a way to retrieve the missed readings later.
• Gateway-side buffering with a backup connection, such as Ethernet plus 4G failover.

Ask vendors exactly how missed data is recovered, and test it during your pilot by unplugging the gateway for an hour.

Coverage matters as much as recovery. Plan gateway positions with the LoRaWAN Gateway and Sensor Placement Guide, and verify links on site with LoRaWAN Link Check.

9. Calibration Certificates and Compliance

“Factory calibrated” is not the same as a calibration certificate. If auditors rely on your readings, ask for traceable calibration per serial number.

Food businesses facing FSSAI or customer audits, pharmaceutical distributors following GDP, and laboratories all need evidence that each sensor reads correctly. In India, that evidence usually means a certificate from an NABL-accredited (ISO/IEC 17025) calibration laboratory, issued per serial number, with a defined recalibration interval (often annual).

Ask these questions before you buy:

1. Is a certificate available for each serial number, and from which lab?
2. At which temperature points was the sensor checked?
3. What does recalibration cost, and does the sensor have to leave site?

Even without a formal requirement, check every pilot sensor against a reference thermometer at your working temperature before you trust it.

Which Is the Best LoRaWAN Temperature Sensor for Your Application?

There is no single best LoRaWAN temperature sensor, only the best fit for your environment, accuracy target and compliance needs. Use this matrix to narrow your shortlist:

Once you know the specification, compare models side by side. Our LoRaWAN temperature and humidity sensor comparison sets the MacSync LTS X1 against the Milesight EM300-TH, Dragino LHT65N and SenseCAP S2101 using each maker’s published figures.

Total Cost: Sensor, Gateway and Network Server

The sensor’s unit price is usually the smallest part of the system cost. Budget for:

• Gateways: one well-placed indoor gateway often covers a building, but cold rooms and steel structures may need more. Size coverage with the Gateway & Coverage Calculator.
• Network server: self-hosted ChirpStack is free software but needs a server and someone to run it. A managed service has a monthly fee.
• Dashboard and alerting: software to store readings, send alarms and generate reports.
• Calibration and recalibration: for regulated sites.
• Battery replacement: factor this in at your real reporting interval.

A cheap sensor that needs a second gateway, or loses data during backhaul outages, is not cheap.

Run a Pilot Before You Buy at Scale

A two-to-four-week pilot with three to five sensors in your worst locations will tell you more than any datasheet. Measure:

• Packet delivery rate (aim for 98% or higher), plus RSSI and SNR at each location
• Spreading factor the network settles on
• Accuracy against a reference thermometer at your working temperature
• Humidity recovery after door-open events
• Data recovery after a deliberate gateway outage
• Battery voltage trend over the pilot period

Macnman offers a pilot kit for exactly this.

Where Macnman Fits, and Where It Doesn’t

Macnman designs and manufactures LoRaWAN sensors in India for the IN865 band:

• MacSync LTS X1 industrial temperature and humidity sensor for warehouses, cold chain, agriculture and outdoor sites
• Narmada TSX1, a compact indoor model
• MacSync LRTD X1 with PT100/PT1000 probes for freezers and process temperatures

The LTS X1 is WPC compliant, comes in battery and 9–36 V DC versions, is IP65 as standard with an IP67 option, and is configured over Bluetooth with the Maya app.

Where the LTS X1 is not the right choice:

• Below −30 °C: its sensor body is not rated for that. Use a probe or the RTD sensor instead.
• Continuously condensing rooms: it is specified for non-condensing conditions.
• On-device display: if you need readings shown on the sensor itself, it has a status LED only.

To get a recommendation for your site, request a quote or speak to a product engineer.

Frequently Asked Questions

What is the best LoRaWAN temperature sensor?
There is no single best LoRaWAN temperature sensor. The right one depends on your environment, accuracy requirement and compliance needs. A compact indoor unit suits offices and server rooms. An industrial IP65–IP67 unit suits warehouses and cold rooms. A probe or RTD sensor suits freezers and process lines. Shortlist by these requirements first, then compare models.

What accuracy do I need for a cold room or pharmaceutical store?
For food cold rooms and 2–8 °C pharmaceutical storage, aim for ±0.5 °C or better at your actual operating temperature, or whatever your quality procedure specifies. Check the datasheet’s accuracy for the sub-10 °C range specifically, because many sensors are less accurate below 0 °C than at room temperature.

Can a LoRaWAN temperature sensor work inside a freezer?
Yes, but the best setup keeps the sensor body and battery outside the freezer and passes an external probe inside. Batteries lose capacity in deep cold, and metal freezer walls weaken the radio signal. Below about −30 °C, use a PT100 or PT1000 RTD probe.

Why does my humidity sensor read 100% in the cold room?
It is most likely condensation on the humidity element, caused by warm, humid air entering through doors or by defrost cycles. Mount the sensor away from doors and evaporator fans, face the opening downwards, and use a filter-protected probe in persistently wet rooms. Most standard sensors are rated for non-condensing conditions only.

Will a LoRaWAN sensor bought for Europe work in India?
Usually not as supplied. European units run the EU868 channel plan, while Indian LoRaWAN networks use IN865 within India’s 865–868 MHz licence-exempt band. You need the IN865 version of the sensor and WPC Equipment Type Approval for that model.

What is the difference between a LoRa and a LoRaWAN temperature sensor?
A LoRa sensor may use a proprietary link that only works with its maker’s own receiver. A LoRaWAN sensor follows the open LoRaWAN standard and works with any compatible gateway and network server, such as ChirpStack or The Things Stack. Check the datasheet for a LoRaWAN version number and a published payload decoder.

How long does a LoRaWAN temperature sensor battery last?
It depends mainly on the reporting interval. A sensor with a 19 Ah Li-SOCl₂ battery may last about 2 years at 5-minute reporting and 7 years or more at 60-minute reporting. Poor gateway coverage, confirmed uplinks and low temperatures shorten this. Always ask what interval a battery claim assumes.

Do I need my own LoRaWAN gateway?
In most Indian industrial sites, yes. Public LoRaWAN coverage is limited, so most deployments install one or more private gateways and connect them to a network server. One indoor gateway often covers a single building, while large or steel-heavy sites may need several.

What happens to the data if the gateway goes offline?
By default, readings sent while the gateway is offline are lost. To avoid gaps, choose sensors that log data locally and let you retrieve missed readings, and gateways that buffer packets and have backup connectivity such as 4G failover.

Do LoRaWAN sensors come with a calibration certificate?
Most are factory calibrated, but a formal certificate is often optional. For audited applications in India, ask for a certificate from an NABL-accredited (ISO/IEC 17025) laboratory per serial number, and plan for periodic recalibration.