The most costly mistake occurs during commissioning

Why a lack of acceptance tests leads to time-consuming service call-outs

With 25 million ‘homes passed’, the roll-out of fibre-optic networks in Germany has long since reached a scale where, alongside civil engineering, route planning and new household connections, the focus is now primarily on operational reliability. But how can the growing dissatisfaction amongst customers, caused by reduced bandwidths or sporadic disconnections, be permanently alleviated?

According to the Gigabit Land Register, by August 2025 over 50 per cent of all households were already able to use or have a fibre-optic connection installed. By early 2026, this corresponded to approximately 14 million households within reach (‘Homes Connected’), more than half of which were active (‘Homes Activated’). In the next 12 months alone, a further 1.5 million households will be activated, and the trend is upwards.

Technologically, GPON
currently dominates with 2.5 Gbit/s downstream and 1.25 Gbit/s upstream, whilst XGS-PON
, offering symmetrical speeds of 10 Gbit/s, will be introduced in many areas as an upgrade over the same fibre. As the number of activated connections rises, an increasing number of issues are becoming apparent in the field: reduced bandwidth, sporadic disconnections, IPTV or OTT disruptions, VoIP problems, or customer complaints because the promised performance is not being delivered. Typical causes often lie in the last few metres: dirty connectors, faulty splices, excessive attenuation, incorrectly assigned PON branches, or misconfigured ONTs. Or a connection that was never fully commissioned.

Optical Fault Locator in an ARGUS F240, with measurement and detection right up to the nearest 1:16 splitter. Measurements are taken at 1625 nm, a maintenance wavelength that can be used whilst the system is in operation.

Acceptance testing is crucial

In PON networks, it is not enough during commissioning simply to verify that there is ‘light on the fibre’. An activated connection is by no means the same as a properly installed one. Anyone wishing to carry out a robust acceptance test today needs to know more than just the optical level.

First, there are the basics: Are the fibre end faces clean? Are the optical levels correct? Is the technician on the correct branch? Is the measured attenuation plausible? Are there any notable events along the route? Is the service at the connection functioning with the promised performance?

Particularly in Germany, with its many end-to-end spliced routes, long home connection chains, distribution points documented in varying ways, and coexisting GPON and XGS-PON structures, these questions are part of everyday practice. The acceptance test thus forms the basis for subsequent operation. Monitoring systems highlight problems, but do not prevent them. The quality of a connection is therefore determined at the time of acceptance.

The fault lies at the fibre end face

Every fibre-optic measurement begins at the fibre end face. Even the tiniest particles or scratches in the core area can lead to significant losses. Test equipment should therefore support fibre inspection. End faces can be inspected and assessed using a video microscope. The cleanliness of the connector determines whether further measurements are reliable. A contaminated end face distorts signal levels and subsequently leads to reflections, unnecessary attenuation and the need for fault-finding. Documenting this condition as part of the acceptance test provides verifiable proof of quality, serves as a reference during subsequent operation and reduces disputes in the event of a fault.

Selective OPM

The second step involves the optical evaluation of the active PON connection. This is where the difference between a standard power metre and a measuring instrument suitable for GPON and XGS-PON becomes apparent. In modern PON networks, several wavelengths are transmitted simultaneously over the same fibre. It is not sufficient to determine an overall level. What is crucial is the filtered measurement of the individual wavelengths. In the downstream direction, this applies in particular to 1,490 nm (GPON) and 1,577 nm (XGS-PON). In networks with video overlay, the 1,550 nm wavelength may also be relevant. The upstream wavelengths of 1,310 nm (GPON) and 1,270 nm (XGS-PON) can only be measured under real operating conditions. An ONT only transmits when prompted to do so by the OLT. Measurement is therefore only possible in through mode, which allows up to five optical levels to be measured bidirectionally at the same time.

Important information can already be obtained here, such as the PON ID. If required, a PLOAM monitor provides the ONU ID or the serial number of an ONT connected to the branch. Through mode is absolutely essential for detecting interference sources such as alien ONTs. In practice, this means: The technician can not only detect whether a signal is present, but also identify which technologies are present at what levels, whether these are sufficient, and how the connection will subsequently behave within the network context. During service calls, this differentiation enables an initial assessment of whether the cause lies within the PON segment or in other areas.

A good signal level is not enough

A conventional power metre shows whether a signal level falls within the specified limits – not how it is achieved. A connection may appear normal yet still contain undetected weaknesses, particularly over short distances with low total attenuation. Such faults become costly later on: initially, they lead to reduced power reserves. As utilisation increases, bandwidths rise or additional services are introduced, these issues turn into faults that can only be localised in the field with considerable effort.

This is precisely where the Optical Fault Locator (OFL) comes into play. Technically, the measurement is based on the OTDR
method, but it is specifically designed for use in PON networks. Measurements are taken at 1,625 nm, a maintenance wavelength that can be used whilst the network is operational. This allows typical sections of fibre to be analysed quickly.

The difference lies in the presentation: instead of a complex curve, the technician receives a guided event analysis along the fibre. Connectors, splices, reflections and end points are automatically detected and displayed along with their distance and insertion loss. This allows the attenuation of the installed link up to the next splitter to be clearly determined and documented. The fully automated measurement thus provides a directly usable service report that requires no specialist knowledge.

An OFL is no substitute for a full-featured OTDR
for detailed analysis. However, for assessing connection runs in the last mile, it provides the information required in the field. Often underestimated, but crucial: the OFL measurement must not be made dependent on the result of the level measurement. It should be an integral part of the acceptance test – regardless of the attenuation budget. This is because only the OFL measurement reveals whether the link has been properly installed or whether weaknesses have already crept in during installation. It is therefore a mandatory component of the connection acceptance test.

In the event of faults, the OFL enables the rapid detection and localisation of typical faults such as poor connections, conspicuous reflections or localised events along the link. This makes it an effective troubleshooting tool in the field. The OFL can also be used to analyse the link and check for installation faults even without an active OLT. Alongside the traditional OLTS measurement using an OPM and OLS, it is ideal for ensuring the physical quality of the link prior to activation.

Flowchart of a connection acceptance test using the Optical Fault Locator and ARGUS F240. This allows the attenuation of the installed link up to the next splitter to be clearly determined and documented.

Testing the connection in practice

The crucial factor is whether the service functions under real-world conditions. A comprehensive acceptance test must therefore be carried out at the protocol level and test the connection in the way it will subsequently be used. This includes simulating a real GPON or XGS-PON ONT with correct access data, serial number and configuration. It is precisely here that errors arise in practice which remain undetected in pure fibre measurements but often lead to disruptions and impair customer satisfaction.

In addition to mere availability, actual performance is crucial. Simply verifying that a connection exists is not enough – the promised performance must also be achieved. That is why speed tests of up to 10 Gbit/s carried out directly at the connection point are a key part of the assessment. This is the only way to conclusively determine whether the interplay between infrastructure, configuration and end devices is functioning correctly.

Furthermore, typical services must be tested: a VoIP test shows whether voice services are provided to a sufficient standard. IPTV services are sensitive to packet loss, jitter and latency. And it is precisely the increasingly important OTT services that often reveal problems first, as they replicate real-world usage scenarios. Many faults do not originate in the fibre-optic cable itself, but at the protocol or configuration level. The connection works from a technical point of view, but services behave erratically, drop out or fail to deliver the expected quality. Without appropriate testing, such faults often remain undetected until the customer reports them.

Another critical point is the home network. From the customer’s perspective, connection quality is assessed on the basis of the WLAN. Whilst modern standards such as WiFi7 enable high data rates, they also reveal when the access point, router and end device are not optimally coordinated. A simple check of the local network can help to clearly distinguish whether the cause of a fault lies with the connection or the home network.

Usability and trust

In addition to measurement technology, usability in the field is a decisive factor in day-to-day operations. Measurement procedures are of little help if they are not understandable or reproducible when under time pressure. Clear operating logic, straightforward procedures, short boot times and a consistent presentation of results lower the barrier to entry and significantly increase acceptance amongst service technicians.

An acceptance test is only reliable if its results are recorded in full and in a traceable manner. Structured logs that can be transferred and processed directly provide a reliable basis for subsequent operation.

Trust is also a crucial factor. Origin, serviceability and data security are playing an increasingly important role. Measurement technology from Europe is often closely aligned with the network architectures in use, offers short support channels and long-term further development.

The solution must work in the field

Ultimately, it is not the individual measurement function that matters, but the ability to arrive at a clear conclusion quickly in the field. Anyone who considers fibre, protocol and service separately will have to look for problems later during operation. Anyone who tests everything as part of an end-to-end process will identify them as early as the acceptance stage. Key factors here are: clear results rather than interpretation, reproducible procedures rather than individual measurements, and comprehensive documentation as the basis for operation.

Integrated test systems such as the ARGUS F240 from intec GmbH demonstrate that this approach can already be put into practice today – developed and manufactured in Germany. For applications without advanced service testing, the ARGUS F200 offers a more compact solution.