Installation & Integration

Communication between the Clipr, Flexr and TMV- sensors and the Gateway Hub is conducted via the LoRaWAN protocol. There are several protocols capable of sending small data packages over long distances using only battery power. Factorylab specializes in LoRaWAN.

Other known protocols are:
➢ LoRaWAN: a world standard wireless data (machine to machine) protocol. Open source, no license needed to broadcast
➢ Sigfox: like LoRaWAN but far less used
➢ NB-IoT: uses e-SIM technology. Ideal for sites with only a few (< 5 pcs) sensors. No need for a gateway but telecom companies charge relatively high fees (± £ 2 per sensor per month only for data)

Customers might expect some kind of Bluetooth or WiFi connectivity but we don’t use that. Bluetooth hasn’t got the range and WiFi needs plugging into a mains electrical connection.
Yes. A comprehensive installation-requirements checklist is supplied during onboarding or at the quotation stage. The better an installation is prepared the smoother it will be.

Typical points of interest are. Availability of:
- Technical drawings, installation lay-out
- Risk analysis
- Asset register
Yes. Detailed location and antenna-placement guidelines are provided in the installation manual and through training. Some basic guidelines are:
➢ Place the gateway in the centre of the site/building
➢ Ideally the gateway should be one floor higher than the majority of the sensors
➢ Technical- plantrooms or MER/SER-rooms are obvious choices
➢ Try not to place in basement of behind thick concrete walls
➢ Mark a gateway with reference or telephone numbers so engineers don’t remove it
Devices are shipped pre-assembled and ready for immediate commissioning, requiring only minimal on-site work. Issues typically arise from poor work preparation and incorrect placement, e.g.
➢ twisting/moving the sensor after it has been clicked on a pipe can easily damage the sensor probe (1 out of 100 goes wrong on average);
➢ missing technical drawings, not knowing what asset or asset type the sensor is supposed to be installed on;
➢ hot water pipes insulated, insulation has to be partially removed➢ inability to connect sensor and gateway. Basic familiarity with radio-equipment installation mitigates
this.

FactoryLab remains available to provide support and training. We also offer training videos and on-site training for engineers.
WIP
Yes they can, depending on the (courtesy of the) gateway owner. A gateway typically sends sensor payloads to a designated LoRaWAN cloud server. From their payloads are forwarded to corresponding receiving parties and databases. When a third party LoRaWAN server forwards Factorylab payloads to the Factorylab database, then the desired connection has been established.

Though technically possible we rarely see this happening. It’s often easier (and cheaper) to just install an own (low cost) gateway and maintain independent system security.
The Gateway Hub maintains a live connection and forwards data to the cloud immediately upon receipt of each LoRaWAN message. Data is forwarded and handled by a dedicated (cloud) LoRaWAN-server.

Factorylab uses AWS-servers physically located in Ireland.
Under line-of-sight conditions, the Gateway Hub can communicate with sensors up to approximately one thousand (1000) metres.

In indoor installations, typical coverage spans roughly three (3) floors or two-hundred (200) metres, depending on building construction and interference.

Example 1: we use one (1) outdoor and one (1) indoor gateway to connect 150 sensors at a mental health care site with buildings max. two floors high.

Example 2: we use only one (1) outdoor gateway to connect 100 sensors at a nursing home with two buildings eight (8) and four (4) floors high. The most challenging is heat reflecting glass, absorbing radio signals.
The Gateway Hub utilises 3G and 4G cellular networks for backhaul communications and operates on the 868 MHz LoRaWAN band for sensor links. This band is license free to use but requires users to comply to the LoRaWAN standard. It is - for example - not allowed to send data every second because this could hinder radio signals from other users. In different parts of the world, different frequencies apply. USA for example uses 915 MHz.

It is important to have cell coverage (4G/5G) for the gateway to be able to send its data into a cloud server!
Yes. Detailed location and antenna-placement guidelines are provided in the installation manual and through training. Some basic guidelines are:
➢ Place the gateway in the centre of the site/building
➢ Ideally the gateway should be one floor higher than the majority of the sensors
➢ Technical- plantrooms or MER/SER-rooms are obvious choices
➢ Try not to place in basement of behind thick concrete walls
➢ Mark a gateway with reference or telephone numbers so engineers don’t remove it
Third party platforms
Integration with third-party platforms is one of our strong points. We distinguish
ourselves from trade companies by offering extended API support. We currently have
working connections with Zetasafe, Legionelladossier and L8MS. API-documentation can be obtained from api.factorylab.nl.

Third party sensors
Any sensor that uses the LoRaWAN protocol can be connected to our gateways and integrated into our platform and from there to other platforms using the API. Factorylab specialises in drinking water installations for which we design our own sensor. For other applications we regularly source third-party sensors.
The standard FactoryLab sensors are not ATEX-compliant but we are happy to source and install compatible third-party ATEX-compliant LoRaWAN sensors on request.

Device Specifications

What is the accuracy and tolerance level (e.g., ±0.5°C)

Our sensors use NTC probes made by Tasseron (www.tasseron.nl). The accuracy of a NTC-probe is +/- 0.2 °Celsius. The electronics and algorithms on the embedded PCB-board add an extra inaccuracy of +/- 0.2 °Celsius giving the end-device an accuracy of +/- 0.4 0.2 °Celsius.
Normal operating ranges on our ClipR/FlexR/TMVR sensors are:
➢ Continuous exposure 1°C to 72°C;
➢ up to four (4) hours per day at 75°C;
➢ up to one (1) hour at 80°C.

The actual NTC-sensor probe itself can withstand temperatures well above 100°C. It’s the PCB, electronic components on it and the ABS-housing of the sensor that limit continuous use in high temperatures.
➢ The Clipr sensor records 4 readings per hour and sends its data every hour. That gives 96 temperature readings per day (4 readings x 24 hours)
➢ The Flexr sensor idem like ClipR.
➢ The TMV Monitor records approximately 1440 readings per day.

Interval settings can be altered by FactoryLab, typically for large-batch requests
The ClipR, FlexR and TMV-sensor all feature user replaceable batteries accessible via a sliding panel or opening up the housing, replacing the battery and then closing the sensor again. The whole process takes about 30 seconds for a trained engineer. However allow time to find a sensor, make it accessible and leave the workplace after battery replacement. We calculate with 25-35 sensors a day an engineer can maintain. Please allow for maintenance and calibration every three (3) years.

Detailed steps are provided in each device’s manual and can be embedded in reactive-maintenance plans.
No measurable drift has been identified to date for the Clipr-, Flexr or TMV-sensor.
Interval settings for the ClipR, FlexR and TMV-sensor can be altered by FactoryLab only, typically for large-batch requests.

We have tens of thousands of ClipR’s in the field and in order to maintain consistency and predictability all operation software is embedded and cannot be altered from the outside.
Sampling frequency depends on device type and battery-life constraints. Across the ClipR, FlexR and TMV-sensor, the maximum supported sampling rate is one reading every five (5) minutes.

Higher rates shorten battery life proportionally. The minimum (standard) sampling rate (embedded on ClipR and FlexR) is set to fifteen (15) minutes.
All devices are back traceable to production dates and batches. FactoryLab has not documented failure-rate or MTBF data for their sensors but can draw knowledge from historical usage. A sensor contains no moving parts so when properly maintained we expect the MTBF to be > 10 years. Although not showing defects we recommend however replacing sensors every 6-10 years with the latest hard- and firmware.
The ClipR, FlexR and TMV-sensor all feature user replaceable batteries accessible via a sliding panel or opening up the housing, replacing the battery and then closing the sensor again.

The whole process takes about 30 seconds for a trained engineer. However allow time to find a sensor, make it accessible and leave the workplace after battery replacement. We calculate with 25-35 sensors a day an engineer can maintain. Please allow for maintenance and calibration every three (3) years.

Detailed steps are provided in each device’s manual and can be embedded in reactive-maintenance plans.
Batteries used in Factorylab sensors are from www.saft.com. The minimum life expectation of a battery is 3 years when the sensor is sending data to the gateway at SF-10.

The technology used inside the battery is referred to as Primary lithium, Li-SOCl2.

ClipR and FlexR use LS14250 type battery.

TMV-sensors have a custom made battery pack on board.
All sensors are equipped with a NTC thermistor probe with a B-value of 3977K with a maximum delta of ± 0.8°C at 60°C. This is the standard inaccuracy of our NTC-probes. Sensors are not (standard) individually calibrated at production.

We do however offer calibrated sensors (with calibration certificate) at an additional charge.
We recommend calibration after three years in use. For efficiency reasons, combine battery replacement with calibration

Each component in the ClipR, FlexR and TMV-sensor is validated during production to confirm conformity with its specifications. Calibration is an extra option and can be done at our factory or on-site.

Factory

➢ 10-50 finished sensors are placed on a copper pipe

➢ Temperature controlled hot water (60 °C) is circulating through the test rig

➢ A calibrated high accuracy thermometer ± 0.2 °C is used as reference measurement

➢ Sensors are activated sending their temperature to the portal

➢ Sensors outside the allowed tolerance ± 1.0 °C are discarded

➢ Calibration logs stating min and max deviation for each sensor are generated

On-site

We do not recommend on-site calibration with a clamp-on reference thermometer due to its inaccuracy. To perform a validated on-site calibration use we advise to use a Factorylab mobile testing rig. That allows for the same calibration procedure as described above.
As a manufacturer we declare that our wireless temperature monitoring devices ClipR, FlexR and TMV-sensor comply with:
➢ DIRECTIVE 2014/53/EU; RADIO EQUIPMENT
➢ DIRECTIVE 2014/32/EU; MEASURING INSTRUMENTS

The devices are marked with the official CE-marking.

A Declaration of Conformity (DoC) is available on request.
Each component in the ClipR, FlexR and TMV-sensor is validated during production to confirm conformity with its specifications. Calibration is an extra option and can be done at our factory or on-site.

Factory
➢ 10-50 finished sensors are placed on a copper pipe
➢ Temperature controlled hot water (60 °C) is circulating through the test rig
➢ A calibrated high accuracy thermometer ± 0.2 °C is used as reference measurement
➢ Sensors are activated sending their temperature to the portal
➢ Sensors outside the allowed tolerance ± 1.0 °C are discarded
➢ Calibration logs stating min and max deviation for each sensor are generated

On-site
We do not recommend on-site calibration with a clamp-on reference thermometer due to its inaccuracy. To perform a validated on-site calibration use we advise to use a Factorylab mobile testing rig. That allows for the same calibration procedure as described above.
Yes. Power-loss alarms for the Gateway Hub can be configured through the online portal. Alarms are usually forwarded via email.

Ordering, Delivery & Service

Devices can be ordered through our webshop: https://webshop.factorylab.co.uk/

Only registered users can login and order through the webshop. Discounts are set automatically. There's no need to pay online; an invoice will be sent later. Please don't forget to fill out the PO-details when finishing your order.
ClipR - FlexR - TMV Sensor, Gateway Hub

All Factorylab made sensors (ClipR, FlexR and TMV sensor) are available from stock. Please allow up to 5 working days to be packed, shipped and delivered on site.

Kerlink gateways (both indoor and outdoor) are also available from stock.

Third party sensors (level, leak-detection, water meters) require up to 2 weeks to be delivered on site.
For the Clipr, Flexr and TMV Monitor, the primary battery is recommended to be replaced every three (3) years, subject to environmental conditions. We recommend on-site calibration every three (3) years.

Please consult the user manual supplied with each device for further details on the applicable maintenance schedule.
A sensor has no moving parts, making it unlikely that it stops unless the battery is empty or someone has touched and damaged it. Downtime is automatically detected and reported via the configurable alarm criteria in the online portal; customers may define thresholds and receive automated notifications accordingly. When an entire site stops transmitting there’s most likely a problem with the gateway e.g. someone has disconnected it from it’s power supply. Please call our helpdesk where our engineers can deep dive into the devices and see what’s actually happening.

Typicalparameters investigated by our helpdesk:
➢ Signal strength (spreading factor)
➢ Percentage of data packages not received

Problems with the signal strength can often be resolved remotely by Factorylab engineers who can increase the strength a device is transmitting with. Mechanical problems (empty battery, NTC-probe damaged) have to be resolved on site preferably by a trained Marlowe engineer.
For the exact SLA terms for our various devices, please check your contract with FactoryLab or inquire through the usual channels.
During normal operating conditions:

Water sensors (Clipr, Flexr, TMV-sensor): operational availability meets or exceeds 99% provided the batteries have enough charge. A data packages loss of <20% is regarded as acceptable. Data packages can be lost in mid-air when they collide with other radio-signal. This is normal behaviour. With a maximum loss of 20% data packages, out of 96 readings still 76 come through.

Gateway Hub: operational availability meets or exceeds 99% provided that the unit is properly connected to a functioning power outlet.

FactoryLab portal: operational availability meets or exceeds 99% subject to the continuous availability of the third-party cloud-hosting service on which the portal is deployed.
The following issues are exceptions to the above uptime guarantees:
➢ Local power outages, building‑engineering works or adverse atmospheric conditions;
➢ Disruptions, congestion or collision of LoRaWAN radio traffic, it being industry best practice that up to 20 % packet loss can occur without materially affecting operational effectiveness;
➢ Failure, misconfiguration or intentional disconnection of customer‑provided infrastructure (e.g., network equipment, cabling); or
➢ Force‑majeure events such as governmental actions or labour disputes.

Web Portal, Reporting & Alerts

The portal provides a range of outputs and reports, with generation frequency adjustable to daily, weekly or monthly.

Extensive customisation of reports is available on request; FactoryLab can demonstrate the full capabilities.
Yes. Summary reports can be downloaded as PDF files for external sharing. They act as a “log book” and can be used as such. They will be the prove customers need to be compliant.
Extensive customisation of reports is available on request; FactoryLab can demonstrate the full capabilities.
User-defined triggers are available for:
➢ temperature,
➢ offline status,
➢ battery level and
➢ supply-voltage conditions.

Thresholds may be set for upper limits, lower limits, in-range requirements and duration-based exceptions. We train engineers how to make the best use of alarms. If configured properly, alarms can be used to detect flushes for example. A very powerful use of it.
Yes, downtime is automatically detected and reported via the configurable alarm criteria in the online portal; customers may define thresholds and receive automated notifications accordingly. When an entire site stops transmitting there’s most likely a problem with the gateway e.g. someone has disconnected it from it’s power supply. Please call our helpdesk where our engineers can deep dive into the devices and see what’s actually happening.

Typical parameters not seen by a user:
➢ Signal strength (spreading factor)
➢ Percentage of data packages not received
Alarm rules may be configured to notify when (a) a sensor stops transmitting, such as upon removal, and/or (b) measured values fall outside the normal range, which may indicate damage or tampering.
Customers may define thresholds and receive automated notifications accordingly.

➢ When a sensor is removed it will keep on sending data as long as it reached the gateway. When a senor is removed from site the signal will be lost and the platform will send an alarm (not because someone has stolen the sensor but because it stopped transmitting)
➢ Damaged sensors often generate minus temperature readings which can be recognised as faulty because minus temperatures are not possible in drinking water installations
WIP