Under pressure, the idea finds its way out

Retrospective · R&D 2024–2026 · Province of Noord-Brabant

For two years Neolook, Ascom and Máxima MC Veldhoven worked on SIR NICU Alarm Intelligence – silent alarming with digital video and AI for the neonatal intensive care NICU. The heart of the project fits in a single sentence: the idea proved stronger than the conditions that tried to stop it. SIR ran into an unusual run of setbacks – construction work that made the live demonstration impossible, leadership changes at both partners, a delayed national care programme. And yet that same period produced finally nine valuable proofpoints, live deployment in six hospitals and two accepted European programmes. That is what happens when an idea is real because it is born at the bedside and comes under pressure: it works its way up, left or right, the way a plant grows up between the tiles of the sidewalk.

Born at the bedside

An intensive-care nurse responds to an average of 150 to 400 alarms per patient per shift. Up to 80 % of them require no action – no clinical urgency, no danger. Yet every alarm must be assessed. When so much time of a nurse’s working time goes to alarm management, priorities bend out of shape – not through inattention, but through the structure of the system itself. Thirty percent of ICU nurses report symptoms of burnout; 10 to 15 % leave the profession early. In the Netherlands a lack of 15% and 20% nurses is foreseen for NICU and PICU.

Alarm fatigue is not a comfort problem. It is not the loud alarms that drive people away, but the endless stream of unfounded signals that erodes judgement. And every false alarm is at the same time a missed opportunity: a nurse responding to noise is not available for the patient who truly needs an intervention.

«Not to generate fewer alarms – but to give nurses the right context, before they have to act.»

What if clinical alarm data and real-time video came together in one system, so a nurse could assess the context of an alarm at a glance – without walking to the bed for every false signal? With that question, our SIR project began in April 2024, co-funded by the Province of Noord-Brabant. Ascom supplies the alarm infrastructure in dozens of hospitals. Máxima MC Veldhoven is one of the leading NICU centres in the Netherlands and, since 2020, Neolook’s first customer.

Three work packages, one goal

The project ran three parallel work packages. The first focused on the technology: an event-capture architecture, a direct API integration with the Ascom platform, an AI data pipeline and synchronisation of video and vital-sign data on a single timeline. The second covered certification – NEN 7510, an MDR compliance plan and preparation for CE marking. The third mapped the human factor: a pilot design at Máxima MC, clinical validation and human-factors research with end users.

The three ran deliberately in parallel, not in sequence. Regulatory requirements drive architecture decisions; user insights drive day to day usability. That same interdependence also made the project vulnerable to disruption from outside – and so, disruption came.

The paradox

The amount of setback in this project turned out roughly equal to the amount of result that came out of it. That sounds like failure, but the opposite is true. The pressure that cut off the planned route forced the idea to find a different exit each time – and those detours became the lasting outcomes. The live hospital demonstration that fell through became an international demo environment. The hospital programme that stalled became a follow-up of clinical adoptions in other institutions. 

The setbacks gave the results. Oddly enough.

The setbacks, named honestly

A retrospective that polishes the difficulties away is worth nothing. So here is what genuinely strained, without damaging any partner – nearly all the causes lay outside Neolook’s sphere of influence.

How the idea found its way anyway

Every blocked route was followed by a detour. When the NICU demo proved impossible, Neolook and Ascom built a full demonstration environment outside the hospital infrastructure – in the Netherlands, in the United Kingdom. 

Economic value creation got delayed, not destroyed. And when the national IZA programme stalled, Neolook – at the request of Máxima MC’s Executive Board and CMIO – led a workshop at the national mProve gathering in Eindhoven, with the direct result of winning the Noordwest Ziekenhuis account in a competitive procurement.

What stands after two years

The technology from this project forms the basis of three productised core components: Event Capture, the Guard add-on and the Open API add-on. Nine proof points are directly traceable to SIR. The architecture is modular – deployable as an integrated system and as a standalone module alongside existing infrastructure.

Internationally the technology landed at Cedars-Sinai in Los Angeles, where infant-driven feeding and general movement assessment is in progress, and in a volume agreement with UMC Utrecht. UMC Groningen uses motion analysis based on the Event Capture component. Apnoea, bradycardia and desaturation carried over from SIR into the European Unite JIP Neodata+ consortium with the Netherlands, Italy and Romania – and pain scores developed there found their way back to Máxima MC as the basis for joint follow-up research.

Two lessons

The first lesson is clinical, not technical. The value of alarm intelligence lies not in eliminating alarms, but in shifting the decision power – from the nurse to the system, before the nurse has to act. That demands an understanding of the workflow, not only of the data.

The second lesson is about the real world. Clinical validation in a running NICU demands infrastructural preparation that is separate from software development: network, construction phasing and IT policy shape the timeline as much, even more as the software code does. The deeper lesson is the paradox itself: a project that stays under pressure and keeps yielding results says something about the strength of the original idea. Bridging great ideas – from bedside to breakthrough.

This project was made possible in part by support from the Province of Noord-Brabant, under the Subsidy Scheme for Economy, Knowledge and Talent Development Noord-Brabant 2022 (reference C2333881). Neolook is grateful to the province for that support.

From bedside to breakthrough. Born inside hospitals, built with clinical partners.