Heavy uncrewed ground carriers for the worst ground in the country — logistics, resupply, casualty carry and sensor recon where wheels sink and trucks turn back. Tracked, battery-electric, built on the same tracked-chassis heritage since 2017, and guided the fleet way: no radio, no GPS.
The ground family has no confirmed performance figures yet. Everything quantified below is a study configuration — explicitly not a spec.
Every Canadian Shield platform has a brochure like this one — same depth, same status discipline, same refusal to publish a number we cannot source. Organized by family below.
CS-110 · Air
No battery. No motor. No receiver. A kinetic carrier launched at ~189 m/s on shop air.
Read the brochure → Download PDF ↓
CS-111 · Air
A capture-head variant on the Goose CS-110 launch line — fouls rotors and entangles the airframe instead of striking it down.
Read the brochure → Download PDF ↓
CS-112 · Air
Aimed, direct-fire, tethered — the net that stays connected to the gun that threw it.
Read the brochure → Download PDF ↓
CS-120 · Air
Goose's launch chain, plus the power Goose leaves out — a tail motor and bow-thruster ring for mid-flight correction.
Read the brochure → Download PDF ↓
CS-130 · Air
A moulded electric aircraft that carries the network — decoy, relay, electronic support. Non-kinetic. Recoverable.
Read the brochure → Download PDF ↓
CS-140 · Air
A big, slow propeller and an IR-biased eye — orbiting instead of hovering, unheard, unseen, unjammed.
Read the brochure → Download PDF ↓
CS-150 · Air
The bird that flies its whole leg without asking a satellite for a name.
Read the brochure → Download PDF ↓
CS-160 · Air
The Air family's smallest body — a tube-launched, motor-cruising scout priced to lose, not to last.
Read the brochure → Download PDF ↓
CS-210 · Ground
Tracked. Battery-electric. 150 kg over muskeg — with nothing to jam.
Read the brochure → Download PDF ↓
CS-220 · Ground
600 kg of freight over bad ground. Payload is the product — mobility is the constraint it is bought within.
Read the brochure → Download PDF ↓
CS-230 · Ground
Eight coordinated carriers. One fibre trunk. One operator. Nothing to jam.
Read the brochure → Download PDF ↓
CS-240 · Ground
A wheeled route-picker with thermal eyes — and a watch mode measured in days.
Read the brochure → Download PDF ↓
CS-250 · Ground
Tracked. Battery-electric. An anchored winch pulls roughly seven times what the tracks alone can.
Read the brochure → Download PDF ↓
CS-260 · Ground
A compact tracked carrier that places self-reacting tools at the work face — the force closes on the structure, not the machine.
Read the brochure → Download PDF ↓
CS-310 · Water
Persistent survey & carrier AUV — one moulded hull that surveys, carries and watches.
Read the brochure → Download PDF ↓
CS-320 · Water
A soft pulsed-jet payload carrier — slower than the current on purpose.
Read the brochure → Download PDF ↓
CS-410 · Underground
A surface plant that pushes a steerable head into ground with no line, no map, no signal.
Read the brochure → Download PDF ↓
CS-411 · Underground
Coiled-tubing bore, fibre-guided — conduit installed, water table mapped, nothing to jam.
Read the brochure → Download PDF ↓
CS-420 · Underground
Recon, mapping and locating in existing tunnels and culverts — with a block-only interpose capability, always human-confirmed.
Read the brochure → Download PDF ↓
CS-430 · Underground
A passive seismic and acoustic node on long-life battery — reporting only over a buried fibre link.
Read the brochure → Download PDF ↓
CS-435 · Underground
A chain of buried Mole nodes on a fibre trunk — detect, localize, cue. Nothing radiated, nothing to jam.
Read the brochure → Download PDF ↓
CS-440 · Underground
Hybrid swim/crawl. Zero visibility. No diver, no radio, no GPS.
Read the brochure → Download PDF ↓
CS-450 · Underground
It maps, it marks, it parks. Cheap enough to send in numbers.
Read the brochure → Download PDF ↓
CS-460 · Underground
Flow-driven, self-propelled or tethered — it rides the line, it doesn't bore one.
Read the brochure → Download PDF ↓
CS-510 · Industrial
A pneumatic seed-pod launcher, retuned to gentleness. Pods down, evidence back.
Read the brochure → Download PDF ↓
CS-520 · Industrial
A small moulded surface craft — the information and marking layer of a spill response.
Read the brochure → Download PDF ↓
CS-530 · Industrial
A contact-tolerant aerial platform sets the hook. A ground-anchored winch does the pulling.
Read the brochure → Download PDF ↓
CS-540 · Industrial
A caged, umbilical-fed washing platform. Water up the line, evidence back down.
Read the brochure → Download PDF ↓
CS-550 · Industrial
An umbilical-fed applicator platform. The pressure stays on the ground; the machine flies the last few hundred millimetres.
Read the brochure → Download PDF ↓
CS-560 · Industrial
Full-width boom, drift-reduction nozzles, a metered rate held across the pass. The broadcast tool for the uniform job.
Read the brochure → Download PDF ↓
CS-561 · Industrial
Machine vision finds the target; only that nozzle opens. Built to cut total product volume per acre.
Read the brochure → Download PDF ↓
CS-570 · Industrial
A trailerable uncrewed dredger for ponds, marinas, lagoons and channels.
Read the brochure → Download PDF ↓
CS-580 · Industrial
An uncrewed collection craft for the debris and spill material that floats.
Read the brochure → Download PDF ↓
CS-590 · Industrial
A tracked, teleoperated manipulator platform for shoreline and land cleanup — reach, load and slope, enforced as one envelope.
Read the brochure → Download PDF ↓
CS-630 · Industrial
An uncrewed multirotor built to find litter, take only litter, and prove it did.
Read the brochure → Download PDF ↓
CS-600 · Industrial
An uncrewed VTOL cargo family — four sizes of the same machine, from toolbox to bulk lift.
Read the brochure → Download PDF ↓
CS-610 · Industrial
A crew-carrying VTOL work machine. Two tiers — PM-8 single-seat, PM-12 two-seat.
Read the brochure → Download PDF ↓
CS-620 · Industrial
An uncrewed coaxial hybrid rotorcraft for long-line log extraction — turn after turn, hour after hour.
Read the brochure → Download PDF ↓
CS-700 · Industrial
A docking and replenishment station that lets a working fleet return, recharge and redeploy without a hand touching it.
Read the brochure → Download PDF ↓
CS-710 · Industrial
A deployable chain of relay nodes — command, telemetry and position augmentation for civil fleets working past the edge of the network.
Read the brochure → Download PDF ↓
CS-C001 · Industrial
Guidance-fibre consumable cassettes — a physical tether option, alongside or instead of radio.
Read the brochure → Download PDF ↓
CS-901 · Defence
A canopy-retarded delivery pod: land on a crumple zone, right itself, open, release the carrier system.
Read the brochure → Download PDF ↓Grizzly CS-210 is a tracked, battery-electric carrier for bad ground — muskeg, snowpack, slope and slash. It anchors the family's claim: tens of kilometres of real bad-ground range, not a highway logistics number — and we would rather publish that sentence than a prettier figure.
A ground carrier has no airworthiness case to argue before a trial can move, and the Grizzly rides on the group's own tracked-chassis heritage — one machine, built in 2017, in continuous field service for nine years — a Development-maturity platform among Roadmap siblings. It is the near-term co-lead with Goose, and the fleet's lead platform for uncrewed ground logistics.
No radio and no GPS in any defence, underground or tethered design: Grizzly-family carriers navigate by preset waypoint, by trailed optical fibre with a human in the loop, or by onboard autonomous SLAM. Radio and GNSS are permitted only on the civil Industrial line — the surface exception, stated explicitly.
Shield Pod air-deploy is a family goal, not a sized requirement — and the arithmetic has already failed at Grizzly scale. We publish that too.
Six ground platforms on one chassis logic. Every one is a carrier for cargo, sensors or tools — never a strike platform — and every study figure below is a working configuration, not a specification. Platform names and CS-designations are working candidates.
A tracked, battery-electric soft-ground mule on the group's tracked-chassis line, in build since 2018. Logistics, resupply, casualty carry and sensor recon over bad ground — the family's Development-maturity anchor.
500 kg gross · 150 kg payload · ~8.2 kPa ground pressure · ~39 km muskeg range at 20 °C
The Grizzly scaled to hauler mass — floats empty, rides the muskeg limit loaded. Bulk hauling and long-leg resupply beyond the Grizzly's load; carrier only, and the thinnest spec card in the roster — almost everything is still to-model, and we say so.
1,600 kg gross · 600 kg payload · ~11.2 kPa loaded · a ~25 km mixed-ground delivery leg with return, collapsing to ~7 km in deep-winter muskeg on an unverified cold derate
The salvo thesis restated for carriers: resilience through cheap numbers, coordinated over fibre and mesh — no radio — in a strictly non-kinetic carrier/sensor scope. The pack's mind must be cheap in watts: compute power is range engineering.
80 kg per unit · 20 kg payload · 8-unit pack at 3 km radius · ≥6-of-8 availability 96.2% at 0.90 per-unit availability — a binomial upper bound that does not account for common-mode failure risk
A wheeled route-picker and the Grizzly's operational pair — scout plus mule. It scouts trafficable ground and reports the rest; it does not claim the muskeg. Perception-limited, not power-limited: sprint only counts when the perception horizon exceeds ~20 m. A 6×6 configuration remains an open design fork, and "Lynx" is a working name pending clearance.
50 kg gross · 10 kg sensor payload · cruise/sprint 5/10 m/s · ~10.2 kPa — harder on the ground than the 500 kg Grizzly
A tracked robotic engineering vehicle that is force-limited, not power-limited — and we state it plainly: bigger motors buy nothing; only weight, cohesion or an anchor buy force. The winch is the real capability multiplier. All ground-engaging tools are steel, machined and welded in-house.
600 kg gross · 1.0 m blade · 20 kN winch line pull (study class) · net drawbar ~2.7–5.4 kN depending on soil — textbook soil values, pending field verification
A compact moulded-polymer UGV that carries, positions and powers mechanical-access tooling only — it cannot out-muscle its own mass, so the tool does the work, not the vehicle. Anything beyond mechanical access is a counsel-first gated fork, not a design. "Wolverine" is a working name pending clearance.
250 kg gross · 60 kg arm + tool module · max static push ~1.7 kN dry (friction assumed, not yet field-verified) — an order of magnitude short of quoted door-failure loads, which is why the powered-tool architecture wins
Every figure above is engineering estimation from the family study packages — working points and study configurations, not specifications. Engineering roles are mechanical carrier scope only; weaponization of any platform is a counsel-first decision, never a product decision.
The ground family runs the fleet's three no-radio guidance tiers in their ground form: preset waypoint, fibre-tethered, autonomous SLAM. There is no ballistic tier, because a ground vehicle is steered or it is parked. No radio and no GPS in any defence or underground design; radio and GNSS are permitted on the civil Industrial line only.
A mission route loaded before the vehicle moves — no link of any kind while it runs. The ground analogue of the barrel-set round: the cheapest tier, and jam-proof because nothing is emitted or received.
The carrier trails an optical fibre back to the operator — command guidance that is unjammable by construction, with a human in the loop for every decision that matters. On the Wolverine, tool actuation is restricted to this tier by a rule that follows from the package itself.
Onboard passive sensing and simultaneous localisation and mapping — no emissions, no link, no tether, no GPS. The most capable tier; autonomous capability ships behind the same counsel-first review the whole fleet applies.
Not a highway logistics claim — tens of kilometres of real bad-ground range, modelled per ground class and per season, with the winter numbers published alongside the summer ones.
The Grizzly drive architecture descends from the group's own tracked-chassis heritage — built in 2017 and in continuous field service for nine years — and its drawing discipline. Development maturity where the siblings are Roadmap, and we label which is which.
The family is sized from the ground up — literally: track pressure against muskeg bearing limits drives the design, which is how the study caught the Lynx pressing harder than the Grizzly and published the finding instead of burying it. The Grizzly's own ~8.2 kPa figure follows from that same study, and the muskeg bearing thresholds it is judged against still need field verification.
The Wolf pack applies the fleet's depth thesis to the ground: many cheap coordinated units over fibre and mesh, no radio, where no single vehicle loss breaks the mission — a strictly non-kinetic carrier and sensor scope.
Moulded polymer bodywork over welded structure, on moulds cut in the owned 5-axis shop. Cheap enough to buy in numbers, simple enough to fix in the field — the division's cost thesis on tracks.
The engineering platforms are force-limited, and we print it: the Beaver's drawbar and the Wolverine's push are published with their soil and friction caveats, and the winch — not a bigger motor — is the multiplier. We publish the physics that says no.
The fleet's first fire-control answer is a buried listening line, not a weapon. The Trembler Line — registered to the Badger underground family, appearing here as a guest system because it wires directly into ground operations — is a picket of buried, fibre-linked seismic nodes that hears digging, footsteps and vehicles through the ground, localizes the source, and cues a response over buried fibre. It is a detection-and-cueing system, not a strike system: no engagement is ever autonomous, and any interposing action requires an explicit human authorization first.
Buried Mole nodes hear digging, tunnelling, footsteps and vehicle movement through the ground itself. Passive seismic sensing on a buried fibre trunk: nothing is emitted, so there is nothing to jam and nothing to find.
The pattern of arrivals across the line resolves a position, a heading and a source class — telling a machine from a footstep from wildlife. Localization accuracy and false-positive discrimination remain open trials questions, not yet answered.
Position, vector and class travel to the operator station over buried fibre only — the fleet's no-radio rule holds through the entire loop. The cue informs a person; it commands nothing, and nothing in the system can move on a cue alone.
Only a human — never the system — authorizes a physical response. Cued carriers then occupy the threat's path and deny the route: block, not strike. There is no autonomous engagement in any tier, any posture or any configuration, because no such branch exists to invoke.
The picket runs a posture ladder with exactly one terminal state — and it is a hold, not a strike. BOXED — NO STRIKE is not a caption on a diagram: it is the executable end state of the whole system, machines holding ground under human authority, with no further branch to escalate to. For an assessor, that is the design fact that matters most, so we put it in capitals.
The picket armed on a learned site baseline, thresholds riding the site's own noise. No dispatch authority is active — by design, nothing in the system can move on a detection alone.
A corroborated detection puts position, vector and class in front of a human, and the authorization hold opens — a pending decision, never a countdown. If the human does nothing, nothing moves.
Only on an explicit human release — given and logged — do responders converge on a computed block point, with the picket still listening behind them. A retraction order is available at every moment of the run.
The route denied, the machines parked in the way, the event archived. The terminal state is hulls holding ground: no strike branch exists anywhere in the system to authorize, so the ladder cannot end anywhere else.
The ground family's role in the picket is logistics and response: node line kits ride any carrier, Moose bulk-carries multi-line kits for long perimeters, and Lynx can shadow a cued vehicle contact — carriers and sensors, never a strike platform. The Trembler Line itself is registered to the Badger underground family; the full system write-up — node spacing models, localization targets and what is still being proven out — lives on the Badger page, and this page deliberately does not restate it. The Trembler Line is an advancing design: nothing has been buried yet.
