PS
Project Supervisor
Scientific Director — Edge AI
EDGE AI · MOROCCO · 2026
Building the Future of Artificial Intelligence in Morocco
MSAT (Moroccan Smart Augmented Things) is an innovative Edge AI platform capable of making intelligent local decisions in Smart Cities, Smart Agriculture, Smart Industry, Smart Healthcare, Smart Mobility, Smart Buildings and Smart Energy.
Edge inference in 8 msPrivacy by designOffline intelligenceHybrid Edge–Cloud reasoning
SCROLL
About MSAT
A national vision for intelligence at the edge
“Our ambition is to transform Morocco into a leader in intelligent Edge AI systems capable of making autonomous decisions while preserving privacy, reducing latency and minimizing Cloud dependency.”
MSAT unifies sensing, embedded computation and hybrid Edge–Cloud reasoning into a single research and industrial platform. Each node perceives its environment, interprets it with compact neural and symbolic models, and acts locally — while the Cloud supervises, learns and refines the collective intelligence of the network.
MSAT
EDGE CORE
Autonomous
Local models decide on device, without waiting for the network.
Private
Sensitive data stays where it is produced, by design.
Instant
Millisecond inference for time-critical operations.
Independent
Full intelligence even when connectivity disappears.
Architecture
From signal to action, in one intelligent loop
An interactive view of the MSAT pipeline. Select any stage to understand its role in the hybrid Edge–Cloud architecture.
Stage 01
Sensors
Distributed IoT sensing captures the physical world continuously — environment, machines, vehicles, patients, energy flows.
Edge
- Handles urgent and simple decisions
Cloud
- Performs deep analysis
- Retrains AI models
- Global supervision
- Complex reasoning
- LLM assistance
Control Plane
One console for a national fleet of intelligence
MSAT Studio streams live telemetry, model health and arbitrated decisions from every edge node — benchmarked against verified national and global indicators.
MSAT Studio · Live
streaming
Connected IoT devices worldwide (billions)
2030 value is the published forecast16.62023
18.82024
21.12025
40.62030
Reference workloads
Vision · MSAT-V282%
Timeseries · MSAT-T64%
Agentic Reasoner47%
Live decision feed
- Smart CityTraffic anomaly resolved at Casablanca node C-17
- HealthVitals drift detected — clinician notified in 6 ms
- EnergyLoad balanced across 42 buildings, −11% peak
- AgricultureIrrigation model retrained on 2.1M samples
Verified indicators
The measurable case for edge intelligence
Every figure below is taken from a public institutional source, with its publisher and publication year.
Zero-egress telemetry
Aggregates are shipped, raw signals stay on the node.
Fleet-wide retraining
Federated updates roll out node-by-node with rollback.
Explainable actions
Every decision carries confidence and its causal trace.
Application Domains
Ten verticals, one augmented platform
The same MSAT node adapts its models, sensors and decision policies to the environment it is deployed in.
And beyond
Every new domain is an extension of the same node — new sensors, new models, the same trusted architecture.
Deployment Map
MSAT across Morocco
Ten reference deployment scenarios, one per region — select a city to see the use case, sensors and expected impact.
Select a city marker 10 deployment nodes
Deployment scenario
Casablanca
Smart Mobility
Adaptive traffic signals, incident detection and multimodal flow orchestration.
Sensors
Radar, cameras, inductive loops, GPS traces
AI capabilities
Multi-object tracking, adaptive signal control, incident classification
Expected impact
Reduced congestion and journey times
Artificial Intelligence
Eight forms of intelligence, one reasoning fabric
MSAT composes complementary AI paradigms rather than relying on a single model — the right reasoning for the right decision.
Reasoning fabric
A living mesh of intelligent nodes
Each point is an MSAT node running its own compact models. Violet cores host the heavier reasoning and federate learning across the mesh, while pulses trace the decisions travelling between them.
Drag to rotate · move your cursor to explore
Neuro Symbolic AI
Neural perception fused with symbolic rules for explainable, verifiable decisions.
Conversational AI
Natural-language access to any node, in Arabic, French and English.
Agentic AI
Autonomous agents that plan, call tools and coordinate across the network.
Predictive AI
Forecasting of failures, demand, yield and consumption before they occur.
Diagnostic AI
Root-cause identification from multi-sensor signatures and history.
Decision AI
Policy arbitration under constraints, risk and confidence thresholds.
Optimization AI
Continuous tuning of energy, routing, scheduling and resource allocation.
Hybrid Edge Cloud AI
One model family, split intelligently between device and Cloud.
Advantages
Engineered for trust and speed
Measured on MSAT reference nodes deployed across pilot environments.
0 ms
Low Latency
Median on-device inference time per decision.
0 %
Privacy by Design
Sensitive data processed locally, never shipped raw.
0.0 %
Offline Intelligence
Node availability without any network connectivity.
0 %
Energy Efficient
Reduction in transmission and Cloud compute energy.
0/7
Real-Time Decisions
Continuous perception and actuation loops.
0-bit
Secure Architecture
Encrypted identity, storage and node-to-Cloud channels.
Project Team
The people behind MSAT
A multidisciplinary team of researchers and engineers building Moroccan Edge AI.
AI
AI Research Lead
Neuro-Symbolic & Agentic AI
EE
Embedded Engineer
Edge Hardware & Firmware
CL
Cloud Architect
Hybrid Edge–Cloud Platform
DS
Data Scientist
Predictive & Diagnostic Models
UX
UX & Systems Designer
Human–AI Interaction
Contact
Let's build the next node together
Researchers, industrial partners and investors — tell us what you would like to explore with MSAT.