• brown-iposs argues that the central barrier between Level 4 and Level 5 network automation is not capability alone, but operator trust — what it calls the ‘confidence gap’.
  • BITÉ’s fast-paced 4G modernisation and 5G expansion made traditional practices untenable, creating the operational and business requirement for this real-world assessment of RAN autonomy.
  • brown‑iposs positions transparency and validation — showing engineers what the automation sees, decides, and changes — as the practical mechanism to de-risk autonomous execution.
  • The CARAT-led project delivers value through automated radio site adjustment, real-time verification of contractor changes, and predictive optimisation aligned to population movement patterns.
  • Beyond technical outcomes, the biggest shift is cultural: engineers move from debating data trust to acting on trusted data to prevent issues.
  • The business case is framed around three pillars: capex efficiency; opex reduction; and revenue protection.

 

Mobile operators are progressing with automation implementations across aspects of their networks, edging up TM Forum’s progress-defining six levels (0–5) of autonomy.

For a number of operators, Level 4 is coming within reach in some areas, indicating a high level of autonomy, including elements of the RAN. But there is a gap between the ultimate aim of fully autonomous Level 5 operations (where decisions are executed fully automatically) with a risk of this becoming a chasm.

Speaking to TelcoTitans, radio network optimiser brown‑iposs believes the goal of full autonomy is achievable without requiring a blind leap of faith. The key, according to the German specialist, is bridging the “confidence gap” through validation and transparency.

This view is supported by evidence from the Baltics, where brown‑iposs has completed a live evaluation of its CARAT API‑based RAN optimisation solution with BITÉ Group — providing tangible proof of a viable way ahead for the next phase of automation in public networks.

Fast‑growing BITÉ: the autonomy imperative

Operating in a regional market that is bucking trends with annual data traffic growth exceeding 25% and significant fixed-wireless access, multi-country BITÉ Group is moving fast to improve customer experience and efficiency through substantial network investment.

BITÉ recently agreed the sale of its infrastructure business, TeleTower, to investor Stonepeak, creating the region’s first fully independent towerco. Currently running around 3,000 mast sites in Latvia and Lithuania, BITÉ is upgrading its 4G network and expanding 5G, with TeleTower committed to deploying a further 1,200 base stations.

With at least one base station being modernised or changed every day, manual network management became an unacceptable burden. Weekly performance reviews were looking at history, not the present reality. BITÉ decided to seek an automation partner capable not only of keeping pace with its RAN modernisation but also aligned with a broader strategic shift toward comprehensive data analytics.

This led BITÉ Lithuania and brown‑iposs to implement a validation for AI‑based Level 5 RAN optimisation. The project employed the vendor’s CARAT (Classification and Root-Cause Analyzing Tool) solution, enabling optimisation based purely on network performance and user experience.

Building confidence through validation

A major hurdle for operators aspiring to Level 5 network automation is trust.

“ The difference between Level 4 and Level 5 automation is really just about whether someone is still looking. ”

Bernd Schröder, brown-iposs founder.

Bernd Schröder

Bernd Schröder

To bridge this psychological gap, brown‑iposs focuses on transparency. CARAT operates as an open model, allowing BITÉ’s engineers to see exactly what the automation is seeing and doing. The solution integrates into the operator’s evolving data analytics infrastructure, linking previously siloed information on user behaviour, service use, and network technical performance.

This provided BITÉ with a clear view of the algorithms and decision-making processes. The project also took an iterative approach: analysing results, identifying areas for improvement, and implementing refinements together.

“ You need to be able to show a client what the initial input was, what we found, and what we did, so they can validate it. Without that validation, automation will never reach Level 5.

This meant BITÉ’s engineers could understand, validate, and refine the system’s behaviour. It enabled collaborative problem solving when complex scenarios emerged.”

Schröder.

Proof in action: three key use cases

From a project instigated in the wake of Mobile World Congress 2025, brown‑iposs and BITÉ have moved rapidly. One year later, they are discussing milestones from the successful assessment and the next steps for wider deployment. Automation enabled by CARAT is now delivering value in three primary use cases:

  • Intelligent network realignment for outages: radio sites are automatically adjusted to manage the impact when a neighbouring base station goes down (whether planned or due to failure). This changes the operational reality from ‘outage management’ to proactive ‘intelligent compensation’, maintaining coverage and customer experience without human intervention.
  • Quality improvement by lowering interference: continuous monitoring of user experience metrics is correlated to radio conditions. When degradation is detected, AI-defined improvements automatically adjust antenna parameters to minimise interference.
  • ‘Over-shooter’ detection and prevention: cells with excessive coverage cause interference in neighbouring areas. Signal propagation patterns are automatically analysed with antenna parameters optimised to reduce overlap while maintaining service quality in the intended area.

“ The project with BITÉ demonstrates that autonomous network management is not just theoretically possible, but practically achievable with current technology.”

Schröder.

“Can it?” v “Should it?” — the engineering philosophy

A key consideration for operators is not just whether network optimisation is possible, but whether it delivers genuine value. brown‑iposs advocates a philosophy of “the question ‘can it be optimised?’ should always be followed with, ‘should it be optimised?’”.

“Before making any adjustment, you need to validate its value and impact”, says brown‑iposs consultant Ian Ginn.

This is critical in a complex network environment where changing one parameter can have cascading effects, potentially detrimental. brown‑iposs models the impact of changes to ensure a measurable improvement in areas such as throughput.

“ The challenge is really to maximise the throughput, maximise the customer experience. Customer experience is not only measured by throughput but it is an important measure, because, if you can increase the achievable data rates, that means you have reduced interference and enhanced service quality. ”

Schröder.

Transforming the engineering culture

Ian Ginn

Ian Ginn

Perhaps the most profound impact of the project is on BITÉ’s engineering team.

Automation is often feared as a way to reduce headcount, but brown‑iposs frames it as a transformation that elevates the engineer’s role. Schröder and Ginn describe a ‘washing machine effect’ whereby automation does not stop the activity, but instead changes the nature of it, and in doing so elevates expectations and capabilities.

“ It would be typical for engineers on a Friday afternoon to sit together and debate the available network data: whether they have enough and how much it could be trusted. Those conversations have changed: now it’s a question of ‘how can we act to make the network performance and experience better?’ — because the data is there, and it is trusted. ” 

Ginn.

“The value of the conversation is changing, and the value of engineers’ jobs is changing”, Ginn adds. “The focus has shifted from responding to problems to preventing them”.

Extracting business value from the autonomous RAN

Beyond technical achievement, the collaboration is focused on extracting tangible business value across three pillars: capital efficiency; operational savings; and revenue protection.

  • Customer experience and revenue — dynamic network configuration enables operators to be far more responsive to population movements, like the weekend coastal migration. “When you can see and understand these patterns, you can modify the network to reflect this”, says Schröder. By maintaining quality of experience during outages or surges, operators reduce the risk of churn.
  • Opex reduction — operational efficiency is driven by reducing the need for physical interventions, including reducing the need for costly radio site revisits through the ability to validate contractor work remotely. Intelligent automated site adjustment enables smarter energy management, such as powering down assets during low-traffic periods in the confidence that neighbouring cells will cover the gap.
  • Smarter capex — investment is a critical element of modernisation. Schröder notes that greater value can be extracted by optimising existing assets so that investment cycles can be maximised. “You don’t want to put up a site only to realise six months later you need double capacity”, says Ginn. With better data, BITÉ can plan investment more strategically, targeting new sites and hardware only when and where truly required.

Future horizons: Latvia, 5G, and beyond

The success of the initial work has laid the foundation for expansion.

brown‑iposs and BITÉ are now looking to replicate the success in Latvia, where the network shares an identical configuration and vendor landscape with Lithuania, and is managed by a single team. The replication is expected to be rapid, with the newer network infrastructure in Latvia allowing for faster optimisation gains.

Technologically, the partners are also exploring extension of automation into the 5G network, including multi-layer coordination between 4G and 5G, and advanced beamforming optimisation. Other new use cases on the roadmap include:

  • Automated event preparation: readying the network for predictable public events like concerts and sports matches.
  • Machine learning-driven installation checks: automatically detect if physical antenna parameters (tilt/azimuth) match the design, further validating contractor quality.
  • Real-time contractor validation: physical modifications made to radio sites by contractors (such as antenna tilt or azimuth change) are reviewed and confirmed in real-time, minimising costly repeat visits.
  • Predictive network optimisation: the network is responsive to shifting usage, with CARAT detecting patterns to enable network reconfiguration readied in advance, rather than reactively after congestion occurs.

The solution’s ability to handle complex environments is also notable. BITÉ operates in a region with unique geopolitical and geographical challenges, including serving border zones. CARAT’s proven ability to maintain stability and performance in these sensitive conditions adds another layer of validation to its capabilities.

brown‑iposs: tier‑1 expertise, agile delivery

brown‑iposs brings deep subject matter expertise to the partnership with BITÉ.

With its track record of delivering optimisation solutions to tier‑1 operators across Europe, and a heritage in private networks, the company is positioned as a specialist ‘knowledge company’. This is reflected in Ginn highlighting that the richest seam of graduates for brown‑iposs recruitment comes from the field of astrophysics.

“ In terms of problem solving, it’s fundamentally the same. With astrophysics, you’re trying to neutralise everything except for that star you’re trying to study in the distance. We’re just working a little closer to home. The methodology is the same, and the skillsets are very similar”

Ginn.

This scientific rigour, combined with a compact, agile team structure, allows brown‑iposs to bring its tier‑1 experience to agile operators like BITÉ without the bureaucratic layers of larger vendors.

“Unification of data is Level 1 of the automation process”, says Schröder, “It’s the foundation of everything we do”.

For other operators watching BITÉ’s progress, the lesson is clear: Level 5 automation is not a futuristic concept. It is a present-day reality for those willing to bridge the confidence gap through transparency, validation, and a focus on business outcomes.

Topics