M2M Tech
Physical AI
Prepared for
Capability statement
M2M Tech Capability Statement for SaskTel
Sovereign Physical AI for connected and critical operations
Extending connected infrastructure into real-time perception, operational intelligence and governed action.
Track record
300+
Deployments
7 yrs
Operational
Backed by
Research & talent
Technology stack
SaskTel · Business Development discussion
Confidential discussion document
July 2026
01 / 17
M2M Tech
Physical AI
Prepared for
Capability statement
M2M Tech Capability Statement for SaskTel
Sovereign Physical AI for connected and critical operations
Extending connected infrastructure into real-time perception, operational intelligence and governed action.
Track record
300+
Deployments
7 yrs
Operational
Backed by
Research & talent
Technology stack
SaskTel · Business Development discussion
Confidential discussion document
July 2026
01 / 17
M2M Tech
Physical AI
Prepared for
Executive alignment
A complementary path from connectivity to intelligent operations
SaskTel strengths
Recipient- Connectivity and private networking
- IoT networks, sensors and devices
- Managed infrastructure
- Cloud and data services
- Cybersecurity
- Local relationships and field support
- Collaboration and innovation ecosystem
M2M strengths
Author- Edge AI and local inference
- Video and sensor intelligence
- Multimodal event correlation
- Operational AI agents
- Human-approved workflows
- Digital twins
- Evidence capture and auditability
- Rapid Physical AI deployment
Potential joint outcome
Connected assets that can understand conditions, surface validated intelligence and support operational decisions.
M2M Tech · Capability Statement · Confidential discussion document
02 / 17
M2M Tech
Physical AI
Prepared for
The operational intelligence gap
Connecting an asset is only the first step
Many organizations already have networks, cameras, sensors and operational systems. The remaining challenge is turning those assets into timely, explainable and operationally useful intelligence.
1
Connect
SaskTel-aligned
2
Collect
SaskTel-aligned
3
Interpret
M2M
4
Decide
M2M
5
Approve
M2M
6
Act
M2M
7
Measure
M2M
SaskTel already delivers stages 1–2. M2M adds stages 3–7 the interpretation, decision and action layer that turns connected assets into operational intelligence.
M2M Tech · Capability Statement · Confidential discussion document
03 / 17
M2M Tech
Physical AI
Prepared for
Where M2M fits
A Physical AI layer across connected infrastructure
SaskTel × M2M · Physical AI reference architecture
M2M Physical AIPhysical environment
Cameras & LiDAR
Sensors & telemetry
Vehicles & machines
M2M AURA · ARC · MEA
Twin · intelligence · edge
AURA
Digital twin
Spatial context, asset and event representation.
AURA
Scenario visualization
Explore live and simulated operational scenes.
ARC
Operational intelligence
AI-assisted investigation and evidence assembly.
ARC
Governed action
Rules, human approval and full audit trail.
MEA
Edge inference
Local processing for latency, resilience and control.
MEA
Device integration
Event generation, store-and-forward, evidence capture.
Enterprise environment
Dashboards & applications
Operator views, executive KPIs, SaskTel or customer cloud.
Work-order & data systems
Tickets, ERP, historians, security and audit platforms.
SaskTel · Connectivity fabric
Provincial reach, urban to remote
Fibre
LTE & LTE-M
Private LTE
LoRaWAN
SD-WAN
IP-VPN
Remote-site
Final architecture would be determined jointly for each customer environment.
M2M Tech · Capability Statement · Confidential discussion document
04 / 17
M2M Tech
Physical AI
Prepared for
M2M capability portfolio
Physical AI capabilities built for real operations
Video intelligence
Detect, classify, search and summarize operational activity.
Sensor and machine intelligence
Correlate signals across equipment, environmental sensors and software.
Edge inference
Process data locally for lower latency, resilience and data control.
Operational agents
Investigate events, assemble evidence and recommend actions.
Human-approved action
Keep consequential physical-world changes under operator control.
Digital twins
Represent assets, environments, events and operational scenarios spatially.
M2M Tech · Capability Statement · Confidential discussion document
05 / 17
M2M Tech
Physical AI
Prepared for
Connected Communities
From connected community data to operational intelligence
Battery-electric transit bus in an urban setting (illustrative). Image supplied by M2M; used for illustration only and not intended to imply any manufacturer, transit-agency, SaskTel or M2M relationship.
Intended outcome
Improve service delivery, reduce operating friction and help municipal teams act on timely, relevant and accurate data.
Representative use cases
- Traffic and congestion intelligence
- Municipal asset monitoring
- Road, curb and intersection condition analysis
- Water and infrastructure anomaly detection
- Public-space and facility safety
- Service-request validation
- Emergency and operations-centre intelligence
- Evidence-backed municipal workflows
M2M Tech · Capability Statement · Confidential discussion document
06 / 17
M2M Tech
Physical AI
Prepared for
Connected Agriculture
Physical AI for connected farms and food operations
Illustrative broadacre wheat harvest near Steptoe, Eastern Washington, USA. Photo: Charles Knowles, CC BY 2.0. Not Saskatchewan and not an M2M customer site.
Representative use cases
- Crop and packhouse visual inspection
- Environmental and water monitoring
- Equipment utilization
- Worker and livestock safety
- Automated quality grading
- Predictive maintenance
- Field and facility digital twins
- Evidence-based production decisions
M2M proof point
M2M's AgriTwin platform combines edge AI, digital twins, vision quality control and operational agents for farms and food-processing environments. More: agritwin.m2mtechconnect.com
M2M Tech · Capability Statement · Confidential discussion document
07 / 17
M2M Tech
Physical AI
Prepared for
Industrial and mining opportunities
A lower-friction path into high-value industrial operations
Potential joint opportunity area
Workers observing heavy equipment at an open-pit mining site (illustrative). Image supplied by M2M; no specific operator, SaskTel or M2M deployment is implied.
Representative use cases
- Worker and lone-worker safety
- Vehicle and mobile-equipment intelligence
- Restricted-zone monitoring
- Conveyor and process inspection
- Predictive maintenance
- Environmental monitoring
- Remote-site operational visibility
- Asset tracking and utilization
- Evidence-backed incident review
Commercial positioning
M2M can begin with a contained operational problem, validate value quickly and expand only after technical and business evidence is established.
M2M Tech · Capability Statement · Confidential discussion document
08 / 17
M2M Tech
Physical AI
Prepared for
Airport and transportation opportunities
Physical AI across passenger, vehicle and asset flows
Expansion opportunities discussed with SaskTel
Commercial aircraft at terminal gates (illustrative airport operations). Image supplied by M2M; no airport, airline, SaskTel or M2M deployment is implied.
Representative use cases
- Baggage exception monitoring
- Landside and curbside congestion
- Airside safety
- Ground-support equipment visibility
- Passenger and queue intelligence
- Perimeter and access-zone monitoring
- Asset and vehicle coordination
- Evidence-backed incident workflows
Status
M2M is engaged in active airport and transportation business-development discussions. These should not be presented as completed deployments.
M2M Tech · Capability Statement · Confidential discussion document
09 / 17
M2M Tech
Physical AI
Prepared for
Edge deployment model
Built to operate where data is generated
M2M edge control enclosure (illustrative): MEA Spark edge processor, dual analog ingestion modules and isolated industrial power distribution. Image supplied by M2M; configuration shown is representative, not a confirmed SaskTel deployment.
Required capabilities
GPU-accelerated inference
Existing-camera integration
Sensor and machine interfaces
Local data processing
Private-network readiness
Store-and-forward resilience
Offline or constrained-network operation
Secure remote monitoring
Role-based access
Evidence retention
Software lifecycle management
On-premises or hybrid architecture
Technology note · M2M solutions may be powered by NVIDIA technology where appropriate. Configurations are sized per site; compatibility with SaskTel products would be validated per project.
M2M Tech · Capability Statement · Confidential discussion document
10 / 17
M2M Tech
Physical AI
Prepared for
Evidence · Ongoing projects
Live edge deployments across sectors
Three current M2M engagements that mirror the operational conditions SaskTel customers face: mobile equipment, plant-floor uptime and low-bandwidth private-network environments.
Twin Berry Farms
Berry Harvester Vision
Agriculture · mobile equipment
Challenge. Reduce yield loss and improve harvesting decisions directly on the machine, in the field.
M2M approach. Machine-mounted edge vision on the harvester with real-time operator feedback; rugged deployment for variable field conditions and intermittent connectivity.
Edge visionOn-machine AIRuggedized
In delivery
Twin Berry Farms
Conveyor Predictive Maintenance
Food processing · production line
Challenge. Reduce unplanned conveyor downtime and give earlier warning of bearing degradation.
M2M approach. Edge monitoring of vibration, temperature and RPM with anomaly detection, alerting and admin/user workflows connected end-to-end.
Predictive maintenanceSensor fusionAlerting
In delivery · Deep dive on next slide →
DeckGuard Vision program
Maritime Deck-Safety Vision
Maritime · dual-use
Challenge. Detect safety-critical shipboard events on low- or intermittent-bandwidth vessels without continuous video streaming.
M2M approach. On-device inference on the MEA edge platform with events-not-video telemetry and store-and-forward resilience.
Edge CVEvents-not-videoStore-and-forward
In delivery
Source: M2M internal project descriptions. Customer names shown for internal SaskTel discussion; external use of Twin Berry Farms and DeckGuard identifiers requires written customer approval.
M2M Tech · Capability Statement · Confidential discussion document
11 / 17
M2M Tech
Physical AI
Prepared for
Ongoing project · Deep dive
Twin Berry Farms Predictive maintenance at the edge
A live M2M deployment turning multi-day emergency stoppages into scheduled maintenance windows on the primary processing conveyor.
Reactive current paradigm
- · Bearing seizes unexpectedly
- · 1–2 days of unplanned line stoppage
- · Emergency repair mid-harvest
Proactive M2M edge paradigm
- · Continuous multi-sensor monitoring
- · Early degradation detection (weeks out)
- · Scheduled maintenance, zero unplanned downtime
Sensing layer
Custom CNC brackets · vibration, temperature, non-invasive Hall-effect RPM at each bearing.
Edge hardware
M2M MEA Spark + dual analog modules in a rugged, fully parallel enclosure.
Local pipeline
On-prem pre-processing of high-frequency signals offline capable, zero cloud dependency.
Application layer
Maintenance UI and alerting workflows translating raw voltages into actionable insight.
Rugged independence
Fully parallel install no interference with OEM equipment, no facility electrical tie-in.
Compound insight
Vibration + temperature + RPM fused locally to predict mechanical stress before thermal escalation.
Protected margins
Shifts a recurring multi-day outage into a controlled, planned service window.
Source: M2M internal delivery record, Twin Berry Farms conveyor project. Customer named for internal SaskTel discussion only.
M2M Tech · Capability Statement · Confidential discussion document
12 / 17
M2M Tech
Physical AI
Prepared for
Twin Berry Farms · Technical detail
Anatomy of a predicted failure
Multi-modal fusion identifies mechanical stress days sometimes weeks before the bearing generates enough heat to fail.
Intervention window: weeks not minutes
Vibration diverges from baseline long before temperature climbs. Alerts fire in Phase 2, giving crews time to schedule the swap.
- 1Normal operationT − monthsBaseline RPM, temperature and vibration signatures logged locally.Vibration flat · Temp ~30 °C
- 2Mechanical stressT − weeksVibration streams detect micro-anomalies. Intervention window opens.Vibration rising · Temp stable
- 3Thermal escalationT − daysFriction drives a pronounced temperature spike final warning.Temp climbing past 80 °C
- 4SeizureT = 0RPM collapses to zero. Avoided in the M2M paradigm via earlier alerts.RPM → 0 · unplanned downtime
Source: M2M Twin Berry Farms deployment. Chart is illustrative of the phase pattern observed; scales vary per site.
M2M Tech · Capability Statement · Confidential discussion document
13 / 17
M2M Tech
Physical AI
Prepared for
Evidence · DeckGuard Vision deep dive
Edge safety vision without continuous video streaming
The closest analog in M2M's portfolio to a SaskTel private-network edge deployment: operates on-vessel, tolerates intermittent connectivity, and emits structured events rather than raw video.
Detection scope
Targeted CV models for shipboard and flight-deck operations
PPE and safety-zone violations
Person-down and man-overboard detection
Cargo and sling-handling anomalies
Near-miss events
Why it matters for SaskTel. The same architecture on-device inference, low-bandwidth event telemetry, store-and-forward maps directly to remote community, mining, agriculture and airport deployments where continuous cloud video is not viable.
Architecture characteristics
On-device inference on the MEA edge platform
Events-not-video telemetry over constrained links
Store-and-forward resilience for offline periods
Uses existing shipboard camera infrastructure
Tuned for low false-alarm rate under motion, weather, lighting
Qualifier
DeckGuard Vision is a dual-use maritime program. External naming, imagery or operational detail beyond this internal deck requires prior written approval from the customer.
Source: M2M DeckGuard Vision project description (internal). No naval customer, vessel operator, SaskTel or third-party endorsement is implied.
M2M Tech · Capability Statement · Confidential discussion document
14 / 17
M2M Tech
Physical AI
Prepared for
Evidence · Delivered outcomes
Recent M2M delivery track record
Five completed engagements across Western Canada, spanning food, energy, buildings and environmental compliance the delivery muscle behind the SaskTel proposal.
ThisFish Inc.
Seafood traceability · BC
99%
fish-weight prediction reliability
Dual-stream CNNs + YOLOv11 for real-time seafood property prediction.
Properate
Home energy · BC
400K+
property records processed at <2s each
HPXML-compliant data pipeline with 92% equipment-label extraction accuracy.
ChmlTech Ltd.
Green hydrogen microgrid · BC
Real-time
fuel-cell and electrolyzer forecasting
ML models + dashboard for microgrid performance and downtime prevention.
The Door Face Panels
Smart entry systems · BC
Split-device
ESP32-C6 + NVIDIA Jetson prototype
Modular smart door with virtualized MATTER testing and AI behaviour models.
KeeWee Solutions
Emissions compliance · AB
Multi-agent
CH4 prediction and regulatory RAG
Lagged time-series pipelines and scalable AI agents for NOx / CO2 expansion.
Customer voice
"The team has successfully delivered all project deliverables. They also explored different experimentation approaches toward the same goal, making it a valuable exercise."
Connor Swain, CTO, ThisFish Inc.
"Deliverables met expectations in code quality and documentation. Everything was well-structured and clear, with timely updates and a commendable level of transparency."
Artur Akhmetgareev, Project Manager, Properate
Source: M2M Manufacturing Sector Case Study (internal). Metrics attributed to M2M's own case-study record; testimonials reproduced verbatim.
M2M Tech · Capability Statement · Confidential discussion document
15 / 17
M2M Tech
Physical AI
Prepared for
Commercial engagement options
Start with the smallest engagement that can prove value
Stage 1
Opportunity and technical discovery
2–4 weeks
- Use-case definition
- Site and data assessment
- Connectivity review
- Technical prerequisites
- KPI and ROI hypothesis
- Pilot recommendation
Stage 2
Rapid demonstrator
6–10 weeks
- Sample or limited live data
- One workflow
- Limited device scope
- Technical validation
- User feedback
- Pilot plan
Stage 3
Operational pilot
≈ 4–6 months
- Live operating environment
- Edge deployment
- Workflow integration
- User validation
- KPI measurement
- Production roadmap
Stage 4
Production rollout
Scope-dependent
- Multi-site deployment
- Support model
- Security hardening
- Lifecycle management
- Additional workflows
- Managed-service options
Commercial note · A typical turnkey proof of concept may be approximately CAD $200,000 over four to six months, depending on hardware, site conditions, data availability, integrations, security requirements and operational scope.
Illustrative planning reference not a fixed package price, quotation or commitment.
Illustrative planning reference not a fixed package price, quotation or commitment.
Delivery principles across every stage
End-user participationSmall iterative releasesTransparent prerequisitesDefined success criteriaSecurity and governance reviewBusiness-value validation before scale
M2M Tech · Capability Statement · Confidential discussion document
16 / 17
M2M Tech
Physical AI
Prepared for
Proposed next steps
Select one opportunity and define the smallest credible pilot
Step 1
Select one SaskTel-sponsored customer opportunity.
Step 2
Conduct a 60–90 minute use-case workshop.
Step 3
Map connectivity, security, data and operational requirements.
Step 4
Confirm the smallest measurable pilot.
Step 5
Prepare a technical demonstration for the customer.
Step 6
Agree on roles, account ownership and commercial structure.
M2M is ready to support SaskTel with a focused technical demonstration, transparent pilot scope and locally supported delivery model.
Priority discussion areas
- Connected Communities and Saskatoon
- Agriculture
- Mining and industrial operations
- Airport and transportation opportunities
- SaskTel 5G Innovation Lab – Powered by Samsung
- SIMSA and regional industry events
M2M Tech · Capability Statement · Confidential discussion document
17 / 17