We turn live data from physical systems into decisions made in time
If you run an operation
You have data and no answers
Solutions & use cases -->
If you are building a consortium
You need a technical partner who ships
R&D track record -->
consortium partners & co-developers · 2009–2026
Organisations we have built, integrated or validated technology alongside inside European research consortia. Full list in the R&D section.
01 – technology
One engine. Eight layers.
Nissatech builds real-time analytics for machines, processes, vehicles and people. One engine — detect what changed, explain why, predict what follows, act before the outcome is fixed. Proven across 22 European research projects and deployed in manufacturing, energy, mobility, health and sport.
L01
Stream ingestion & sensor fusion
EDGE · IOT · MULTI-SOURCE
Heterogeneous sensor data — vibration, energy, air quality, cameras, wearables, telemetry — normalised and fused into one continuous, queryable stream. Non-intrusive by design: external sensing where touching the machine is not an option.
L02
Complex event processing
CEP · PATTERNS · TEMPORAL LOGIC
The founding competence of the company. Patterns that only exist across time and across sources get recognised while the stream is still moving, at volumes where batch processing is useless.
L03
Behaviour & variation detection
ANOMALY · DRIFT · PROCESS-LEVEL
We monitor the behaviour of a process as a whole rather than chasing isolated anomalies. That is what catches the slow, compound drift that individual thresholds never fire on.
L04
Root-cause & impact analysis
EXPLAINABILITY · CAUSAL
A flagged deviation with no explanation gets ignored by operators. Every detection carries its likely cause and its consequence, so the alert can be acted on rather than dismissed.
L05
Proactive prediction
FORECAST · PRE-EMPTIVE ACTION
Forecasting the trajectory of the process and triggering intervention inside the window where it still matters. This is the core of our proactive-sensing work and the reason systems built on it prevent rather than record.
L06
Cognitive & hybrid digital twins
SIMULATION + LEARNING
Physics-based models fused with learned models, so the twin keeps improving from operating data instead of drifting away from the asset it represents.
L07
AI at the edge
ON-DEVICE · LOW LATENCY · 5G
Inference where the data is produced, for the cases where latency, bandwidth or confidentiality make a round trip to the cloud unacceptable.
L08
Conversational & agent layer
LLM · AGENTS · DATA VALUATION
Natural-language access to the analysis for people who are not analysts, plus the emerging work on pricing and contract-based sharing of the data these systems generate.
Engineering services
Software integration
We join a stack that already exists — SCADA, ERP, MES, club systems, third-party sensors — rather than asking for a rebuild.
Engineering services
Data Diagnostic Laboratory
A structured engagement for big-data management: what is in your data, what is usable, and what it can realistically predict.
Engineering services
Mobile & wearable systems
Android, Wear OS, embedded sensing and secure mobile data handling, built for environments where the device gets abused.
02 – solutions
Six domains. The same underlying engine.
Each solution below started as a European research project and was carried through to a working system. The provenance line under every card shows where it came from and how far it has been taken.
Manufacturing
Zero-defect & process quality
Spot unusual process behaviour and validate it before defective product starts coming off the line. Quality monitoring that a mid-size manufacturer can actually afford to deploy.
- CogniZDM — cognition-driven zero-defect manufacturing
- CognEcoQuality — eco-aware quality across energy, emissions and material waste
- Non-intrusive sensing: external energy, air-quality and camera inputs
Origin ProaSense · ZDMP · KITT4SME TRL 7–8
Energy
Predictive maintenance
A new generation of predictive maintenance for assets that are too small to justify a large industrial monitoring contract but too expensive to run to failure.
- Cog4Sol — predictive maintenance for mid-size solar power plants
- Degradation and yield-loss forecasting from operational data
- Delivered with Nissahub, our industrial AI delivery arm
Origin ProaSense · COGNITWIN TRL 6–7
Process industry
Cognitive digital twins
Hybrid twins for continuous processes — aluminium, steel, silicon, energy recovery — where the model has to keep learning from the plant it mirrors.
- Self-learning twins combining simulation and operating data
- Deployed alongside SINTEF, Fraunhofer IOSB, DFKI and Cybernetica
- Asset Administration Shell and streaming-infrastructure integration
Origin COGNITWIN · DIGITbrain TRL 6–7
Data economy
Data valuation & sovereign sharing
Data that sits in one organisation is worth a fraction of what it is worth shared under contract. We work on how to price it, contract it and share it without giving up control.
- Contract-based data sharing and automated negotiation plugins
- Context-aware valuation of personal and industrial data
- Live pilot running on our own production platform
Origin UPCAST (Horizon Europe) TRL 5–6
Mobility & health
Remote monitoring & intelligent mobility
Two long-running lines from the same engine: personalised remote monitoring of people, and real-time fusion of multi-source data for urban mobility.
- MCA — personalised remote monitoring assistant
- Real-time processing of complex transportation events
- Proactive traveller and operator notification
Origin OPTIMUM · FI-STAR TRL 6–8
Sport & human performance
Our most commercially mature branch, and the proof that we take research all the way to market. The same detection-and-prediction engine applied to the human body under load.
- Fun4Fit — live heart-rate zones and a regional club leaderboard
- FitBot, Running Analyzer, SmartMonitoRING, S4FPro
- 100+ partner clubs; professional-sport reconditioning work
Origin ProaSense · FI-STAR · UPCAST pilot TRL 9 · In market
Smart4Fit runs under its own brand at smart4.fit. Industrial delivery runs through Nissahub.
03 – use cases
Where it has actually run.
Solutions describe what we can build. These are deployments — a named environment, a real constraint, and a number that changed.
Manufacturing
Defect prevention on a production line
Automotive component manufacturing · ZDMP platform
Multivariate monitoring of process behaviour rather than isolated sensor thresholds, so compound drift is caught and validated before the run produces scrap.
Pre-defect
Detection window
Eco-aware quality monitoring
Manufacturing SME · KITT4SME cascade
Energy, emissions and material waste measured non-intrusively with external sensors and cameras, then modelled together as one quality signal an SME can act on.
3 wastes
Energy · emission · material
Hybrid digital twin for continuous processes
Aluminium, steel and energy-recovery plants · COGNITWIN
Technology-provider role alongside SINTEF, Fraunhofer IOSB and DFKI, building twins that keep learning from plant data instead of drifting from the asset.
6 plants
European industrial sites
Proactive traffic and traveller intelligence
Urban mobility operators · OPTIMUM
Architecture and engine for real-time fusion of multi-source transport data, turning congestion signals into interventions instead of after-the-fact reports.
Real-time
Multi-source fusion
Load dosing and injury prevention
ŽKK Student & ŽRK Naisa, Niš · Smart4Fit S4FPro
Live heart-rate response used to dose load per athlete across a full season. Both clubs finished national runners-up in 2025/26 with continuity through the season.
−65%
Reported injuries
Member retention through live monitoring
Trening centar Active Life, Odžaci · Fun4Fit
Real-time zone display and gamified progress turned into a retention mechanism; sustained growth forced a move to a larger facility in 2026.
+40%
Member retention
04 – research & development
Twenty-two European projects since 2009.
We take part as software integrator, developer, validation partner, or provider of real use cases from our own operations — and we commercialise the result afterwards. The named entries below are confirmed; the remaining archive entries are listed as Project n pending confirmation.
Role
Technical partner
Architecture, core components and integration work — the part of the consortium that has to produce running software.
Role
Validation & use cases
Real deployments in real environments, including our own commercial platform as a live pilot.
Role
Exploitation
A partner with a track record of turning project results into products that survive after the grant ends.
#
PROJECT
OUR CONTRIBUTION
PERIOD
PROGRAMME
IMAGINE
Innovative End-to-end Management of Dynamic Manufacturing Networks
Real-time monitoring and event processing across dynamic manufacturing networks. FP7 GA 285132.
2011–2014
FP7
FED4FIRE
Federation for FIRE
Federated experimentation facilities for future internet research. FP7 GA 318389.
2012–2016
FP7
FITMAN
Future Internet Technologies for Manufacturing
FIWARE-based trials for smart, digital and virtual factories. FP7 GA 604674.
2013–2015
FP7
FI-STAR
Future Internet Technologies for Manufacturing
Future-internet eHealth pilots; foundation of our remote-monitoring line. FP7 GA 604691.
2013–2015
FP7
ProaSense
The Proactive Sensing Enterprise
Our core technology project: detect, understand and act on complex real-time streams. FP7 GA 612329.
2013–2017
FP7
06
OPTIMUM
Multi-source Big Data Fusion Driven Proactivity for Intelligent Mobility
Architecture and engine for real-time processing of complex transport events. H2020 GA 636160.
2015–2018
H2020
07
BEinCPPS
Business Experiments in Cyber-Physical Production Systems
CPS experiments with European manufacturing SMEs. H2020 GA 680633.
2015–2018
H2020
08
MIDIH
Manufacturing Industry Digital Innovation Hubs
CPS and IoT components plus digital innovation hub services. H2020 GA 767498.
2017–2020
H2020
09
COGNITWIN
Cognitive Factories via Proactive, Self-learning Hybrid Digital Twins
Technology provider for hybrid cognitive digital twins in the process industry.
2019–2023
H2020
10
DIGITbrain
Digital Product Brain / Manufacturing as a Service
Digital-twin memory and cross-border manufacturing experiments. H2020 GA 952071.
2020–2023
H2020
11
CogniZDM
Cognition-driven Zero Defect Manufacturing (ZDMP cascade)
Proactive detection of complex process variation, integrated into the ZDMP platform. GA 825631.
2021–2022
H2020 cascade
12
CognEcoQuality
Cognition-driven eco-aware process quality (KITT4SME cascade)
AI monitoring of energy, emissions and material waste for SME-affordable eco-aware quality.
2022–2023
H2020 cascade
13
SpeedUP! Europe
FIWARE acceleration programme
Early acceleration funding for productisation of research assets.
2015–2016
FIWARE accelerator
14
SOUL-FI
FIWARE acceleration programme
Acceleration funding supporting the move from research to market.
2015–2016
FIWARE accelerator
15
AI REDGIO 5.0
Regions and (E)DIHs alliance for AI-at-the-Edge adoption by Industry 5.0 SMEs
AI at the edge for European manufacturing SMEs. Horizon Europe GA 101092069.
2023–2026
Horizon Europe
16
UPCAST
Universal Platform for Contract-based Data Sharing
Data valuation and contract-based sharing; our own platform runs as a live pilot. GA 101093216.
2023–2026
Horizon Europe
17
Smart4Fit
Real-time human-performance monitoring ecosystem
The commercial platform: wearables, real-time zones, AI analytics, gamification. TRL 9.
2016-present
Own product
18
Fun4Fit Community
Regional competitive fitness network
100+ clubs on one live leaderboard with four competition formats.
2023-present
Own product
19
MCA
Mobile Collaborative Assistant
Personalised remote monitoring for health and wellbeing scenarios.
2016–2019
Own product
20
Cog4Sol
Cognition-driven predictive maintenance of solar power plants
Predictive maintenance for mid-size solar plants, delivered via Nissahub.
2023–present
Own product
21
Data Diagnostic Laboratory
Big data management service line
Structured diagnostic engagement for organisations with data but no answers.
2016–present
Own product
22
Project 22
To be confirmed from the project archive
Details pending confirmation.
-
To confirm
SINTEF
Norway
Fraunhofer
Germany
DFKI
Germany
FZI
Karlsruhe
KIT
Karlsruhe
NTUA
Athens
VTT
Finland
UNINOVA
Portugal
Jožef Stefan Institute
Slovenia
ICCS
Athens
Politecnico di Milano
Italy
TU Berlin
Germany
INRIA
France
imec
Belgium
University of Oulu
Finland
University of Southampton
UK
CERTH
Greece
BIBA
Bremen
MTA SZTAKI
Hungary
UPV
Valencia
Coventry University
UK
University of Innsbruck
Austria
Trinity College Dublin
Ireland
Lancaster University
UK
University of Bristol
UK
University of Niš
Serbia
Hydro Aluminium
Germany
Elkem
Norway
Saarstahl
Germany
Sidenor
Spain
SHI FW Energia
Finland
Aker Solutions
Norway
Siemens
Germany
Volkswagen
Germany
Whirlpool
Italy
ATOS
Spain
Engineering
Italy
TXT e-solutions
Italy
INTRASOFT
Luxembourg
Cybernetica
Norway
TEKNOPAR
Türkiye
Scortex
France
Libelium
Spain
Holonix
Italy
Innovalia
Spain
Softeco Sismat
Italy
Centro Ricerche Fiat
Italy
Noksel
Türkiye
Datapixel
Spain
BT Group
UK
Santer Reply
Italy
CloudBroker
Switzerland
These organisations were co-partners in the projects listed above. Listing them describes the company we have kept inside European research programmes, not a customer relationship.
05 – about
A small team that has been solving one problem for 17 years.
Nissatech Innovation Centre is a research-performing SME in Niš, founded in 2009 by people who came out of the European research system and wanted to build products with it rather than papers about it.
Our advantage is not scale. It is that a dozen engineers have worked on the same class of problem — understanding a live stream well enough to act on it — across factories, hospitals, transport networks, power plants and gyms. Each domain taught the engine something the others could use.
We like to challenge the status quo and build technology that improves how a business actually runs. The well-being and development of the people on the team are goals in themselves, not a means to an end. That is the reason the same group is still here after seventeen years of projects.
2009
Founded in Niš as a research-performing SME by Dr Nenad Stojanović.
2011–2013
First FP7 consortia — IMAGINE, FED4FIRE, FITMAN, FI-STAR. Real-time processing meets manufacturing and eHealth.
2013–2017
ProaSense establishes proactive sensing as the company’s core technology.
2015–2020
Horizon 2020 scale-up — OPTIMUM, BEinCPPS, MIDIH. Mobility and cyber-physical production systems.
2016
Smart4Fit launched: the engine is pointed at the human body and becomes a commercial product.
2019–2023
Cognitive digital twins with COGNITWIN and DIGITbrain; cascade-funded zero-defect and eco-quality work.
2023–2026
Horizon Europe — AI REDGIO 5.0 and UPCAST. AI at the edge, and the economics of data itself.