Applications  ·  By task · by industry

Hard tasks, proven on real lines.

Every clip below is SigmaKit running a task that conventional automation struggles with. No CGI, no cherry-picked lab takes.

01
By task

Sixteen skills the floor
said were impossible.

Handle Deformable Cable

#DeformableCableHandling#BimanualManipulation#VisionGuided#ContactFeedback#CableRelocation
ChallengeThe shape and position of a flexible cable keep changing, and precise bimanual coordination is required to detach it from an existing hook and move it.
CapabilityCarbonSix's bimanual robot control, based on vision and contact feedback, responds to the cable's changes in real time, performing stable grasping, separation, transfer, and repositioning.

Insert & Hook Deformable Cable

#CableInsertion#Hooking#VisionGuided#ContactFeedback#BimanualManipulation
ChallengeThe shape and slack of the cable end are never quite the same, so accurately inserting and hooking it at the target position is prone to repeated error.
CapabilityUsing vision and contact feedback, CarbonSix tracks the cable end's position in real time while both arms coordinate to insert and hook the cable precisely and securely at the target position.

Deformable Cable Handling

#ContactAwareAISkill#ConnectorInsertion#AdaptiveRetry#PreciseForceControl#TaktTimeOptimization
ChallengeFloating cables and compliant connectors introduce uncertainty during alignment and mating. Linear insertion attempts often fail due to misalignment and surface ambiguity.
CapabilityCarbonSix leverages a contact-aware AI skill to detect insertion cues through physical interaction. It maintains controlled contact, adaptively retries when needed, and executes time-optimized motion to reduce takt time while preserving reliability.

Hook Latching

#HybridSkill#AIBasedSkill#RuleBasedSkill#TaktTimeOptimization#ManufacturingAutomation
ChallengeOscillating reflective hooks and positional variance in engine blocks disrupt stable insertion workflows. Pure rule-based control struggles with variability, while AI-only control may not fully optimize cycle time.
CapabilityCarbonSix combines AI-based skills for adaptive perception and alignment with rule-based skills for deterministic high-speed execution. This hybrid skill architecture enables reliable and takt-time-efficient hook latching under dynamic conditions.

Precision Peeling

#ThinFilmHandling#SurfaceConformalMotion#AdaptiveForceControl#PositionVariance#MaterialProcessing
ChallengeWrinkled semi-transparent films and shifting wheel positions create unstable peeling dynamics. Vertical lifting often concentrates stress and increases the risk of tearing.
CapabilityCarbonSix applies surface-conformal motion and adaptive force modulation to distribute stress along the wheel contour. By reflecting human peeling strategies, it achieves clean detachment without damage even under positional variance.

Coordinated Bimanual Manipulation

#BimanualManipulation#DualArmCoordination#FlexiblePartExtraction#MaterialSeparation#ManufacturingAutomation
ChallengeExtracting a deformable timing belt from a transparent pouch requires precise dual-arm coordination. The pouch collapses easily while the belt twists unpredictably, making stable separation difficult for single-arm or fixed automation systems.
CapabilityCarbonSix enables coordinated bimanual manipulation through intelligence-driven synchronization. One arm stabilizes the packaging while the other isolates and transfers the flexible component. The system maintains controlled interaction dynamics throughout separation and placement, ensuring stable downstream processing.

Arrangement

#IrregularPlacement#PositionVariance#FlexibleObjectGrasping
ChallengeDumplings appear in irregular orientations and positions, making it difficult for traditional systems to align them consistently. Fixed automation fails when dumplings vary in shape or layout.
CapabilityOur technology perceives dumplings in unstructured states and intelligently re-orients them. It adapts flexibly to irregular arrangements, enabling accurate alignment before further processing.

Assembly

#PositionVariance#TiltVariance#Retry#Assembly(Insertion)#Electronics
ChallengeIn label maker assembly, the position and tilt change constantly, creating irregular conditions. Insertions often fail on the first attempt, requiring retries.
CapabilityOur technology senses and responds to positional and angular variations in real time. It completes the assembly in a single attempt when possible, or performs additional actions to ensure a reliable result.

Peeling-off

#FlexibleObject#ThinFilm#IrregularPattern#PreciseForceControl#MaterialProcessing#Battery
ChallengeThin, transparent, and fragile films are difficult to detect with traditional vision sensors. Irregular patterns and positional changes make stable handling nearly impossible, while the rigid motions of conventional robots often cause damage.
CapabilityOur technology delicately detaches films without tearing, even under irregular conditions. By mimicking human-like fine force control, it achieves precision handling that minimizes damage and goes beyond traditional robotic performance.

Cable Insertion & Fastening

#FlexibleObjectGrasping#PreciseForceControl#CableFastening#Electronics
ChallengeFlexible and elastic cables are difficult to handle, requiring delicate force control. Even the slightest misalignment can lead to damage.
CapabilityOur technology replicates human dexterity to achieve precise force control, making reliable cable insertion and fastening possible.

Hanging

#FlexibleObjectGrasping#PositionVariance#Hanging#TonBag#Apparel#MaterialProcessing
ChallengeFlexible materials such as bag straps or ton bags shift and sag unpredictably, making them difficult to handle.
CapabilityOur technology accurately recognizes these variations and securely grasps deformable objects, enabling consistent and reliable hanging operations.

Pick & Placement

#ConveyorOperation#Retry#FoodManufacturing#StableHandling
ChallengeOn moving conveyors, dumplings must be grasped and placed precisely. Traditional systems struggled with motion, and failed grasps required repeated attempts.
CapabilityOur technology reliably grasps dumplings from the moving conveyor and gently places them where they need to go. It stays stable while the line keeps running, and if a grasp fails it retries automatically until the pick is completed.

Hooking

#MovingObject#ReflectiveSurface#ThinMaterial#PreciseGrasping#IrregularTask#ManufacturingAutomation
ChallengeHanging thin, reflective, and moving objects with accuracy is extremely difficult for conventional robots. Motion and glare disrupt stable detection, while the slender form makes secure grasping a challenge.
CapabilityOur technology reliably detects and grasps even thin, reflective, and moving objects, placing them precisely where needed. It adapts with human-like finesse to complex and irregular environments.

Removal

#IrregularTask#PositionVariance#ShapeDeformation#Retry#PreciseGrasping#HomeAppliance
ChallengeAir conditioner filters vary in position, angle, and mounting structure depending on the model. Over time, dust buildup and deformation cause unpredictable resistance, making removal difficult. Conventional robots struggle to adapt to such irregular motion and variability.
CapabilityOur technology detects subtle positional and shape variations in real time, adjusting force delicately throughout the removal process. Even when the filter is stuck or misaligned, it intelligently retries and completes the task without damage. It performs with human-like adaptability and precision, even under irregular conditions.

Unpacking

#Unpacking#BimanualManipulation#PackageSeparation#FlexiblePartExtraction#ManufacturingAutomation
ChallengeBags crumple easily and belts twist inside them, so fixed, single-arm-centric automation struggles to separate and transfer them reliably.
CapabilityCarbonSix coordinates both arms in real time to separate packaging from parts reliably. One arm holds the bag to control its deformation while the other precisely grips and removes the belt for separation and stacking. Consistent control is maintained even under complex deformation, ensuring a stable handoff to downstream processes.

Bolt Bin Picking

#BinPicking#ReflectiveSurface#PrecisionGrasping#PositionVariance#ManufacturingAutomation
ChallengeReflective, glossy surfaces make vision recognition difficult, and small parts' position and orientation vary enough to make stable pick & place difficult.
CapabilityVision-based precision recognition and grasping identify each bolt's position and orientation accurately, even in reflective environments, enabling stable bin picking.
02
By industry

Where the variance lives,
we deploy.

IND 01
Electronics

Connector insertion, cable fastening, precision assembly of consumer and industrial electronics.

IND 02
Automotive

Hook latching on engine blocks, timing-belt extraction, and takt-time-critical line tasks.

IND 03
Battery & Energy

Thin-film peeling and delicate material handling for battery manufacturing.

IND 04
Food Manufacturing

Conveyor pick & place and arrangement of deformable, irregular food products.

IND 05
Home Appliance

Filter removal, part handling, and assembly across model variations.

IND 06
Apparel & Materials

Hanging, ton-bag handling, and flexible material processing.

Have a task like these?