Robot-Type Explorer

Strong at

• Decades-proven reliability at high speed and precision
• Runs 24/7 on a fixed, well-defined task
• Huge integrator and spare-parts ecosystem

Honest limits

• Needs safety fencing or certified safeguarding
• Expensive to re-purpose when the product changes
• Integration typically costs as much as the arm itself

Seen in the wild

• Spot and arc welding lines
• Machine tending on CNC cells
• High-speed handling of heavy or hot parts

Strong at

• Blistering speed on light payloads
• Small footprint drops into existing lines
• Lower cost entry into automation

Honest limits

• Light payloads only and limited reach
• Struggle with unstructured or variable parts presentation
• Product changeovers need engineering time

Seen in the wild

• Food and pharma pick-and-place
• Electronics assembly steps
• Kitting and sorting cells

Strong at

• Removes one of the most injury-prone manual jobs
• Predictable ROI on steady product flows
• Handles weights no human should

Honest limits

• Needs consistent case sizes or smart grippers
• Floor space and safeguarding around the cell
• Underused if volumes are low or seasonal

Seen in the wild

• End-of-line palletizing in beverage and FMCG plants
• Bagged goods stacking
• Depalletizing inbound goods

Strong at

• Works alongside people after a risk assessment, often without fencing
• Redeployable across tasks in a small workshop
• The gentlest learning curve in industrial automation

Honest limits

• Collaborative speeds are slow: throughput is a fraction of a fenced arm
• Force limits cap payload and tooling
• "Collaborative" is a risk assessment of the whole cell, never a sticker on the arm

Seen in the wild

• Machine tending in SME workshops
• Screwdriving and light assembly beside operators
• Quality-check camera handling

Strong at

• Answers the global shortage of skilled welders for repetitive seams
• Batch-of-ten economics, not batch-of-ten-thousand
• Frees certified welders for the complex joints

Honest limits

• Still needs a welder to program, fixture and inspect
• Fume extraction, arc screening and PPE do not go away
• Complex geometry and out-of-position welds stay manual

Seen in the wild

• Repetitive fillet welds on frames and brackets
• Small-batch fabrication shops
• Pre-fab welding in construction supply

Strong at

• Deploys in weeks without changing the building
• Fleet software scales from 2 to 200 units
• Cuts the walking that dominates picker time

Honest limits

• Struggles with cluttered, changing aisles and door thresholds
• Fleet software licensing is a real recurring cost
• Someone still loads and unloads at each end

Seen in the wild

• Goods-to-person picking loops
• Lab sample and stores runs in hospitals
• Work-in-progress moves between cells

Strong at

• Runs the boring dock-to-rack lanes around the clock
• Removes the most common category of warehouse vehicle accidents
• Predictable flows automate cleanly

Honest limits

• Slower than a skilled human driver on varied work
• Damaged or non-standard pallets cause interventions
• Site traffic management is a project of its own

Seen in the wild

• Dock to racking shuttles
• Line-side replenishment
• Finished-goods transfer between halls

Strong at

• Fits spaces and tools designed for people, no rebuild
• One platform can in principle rotate across tasks
• Progress is genuinely fast thanks to foundation-model control

Honest limits

• Field capability is far narrower than launch videos suggest
• Uptime, hands and battery life remain hard limits
• Economics only compete on tasks nothing simpler can do

Seen in the wild

• Tote moving and machine tending pilots
• Repetitive logistics tasks in structured zones
• Dull-dirty-distant inspection walks

Strong at

• More stable and affordable than legs for flat sites
• Longer duty cycles between charges
• Same dual-arm manipulation research line as bipeds

Honest limits

• Stairs, kerbs and thresholds end the mission
• Still pilot-stage for most real workloads
• Aisle widths and lift access constrain deployment

Seen in the wild

• Order picking pilots in e-commerce
• Room service and hospital delivery trials
• Reception and wayfinding pilots

Strong at

• Replaces working-at-height for visual checks
• Minutes to deploy, same-day reporting
• Thermal finds problems eyes cannot

Honest limits

• Weather and wind ground it
• Aviation rules apply: licensing, no-fly zones, insurance
• Data is only useful with a disciplined review workflow

Seen in the wild

• Roof, facade and solar-farm inspections
• Port and construction progress monitoring
• Coastal and agricultural surveys

Strong at

• Days of ground survey become one flight
• Repeatable flights make change visible over time
• Stockpile volumes without climbing them

Honest limits

• Photogrammetry needs processing power and skill
• Accuracy claims depend on ground control discipline
• Same aviation rules and weather limits as any drone

Seen in the wild

• Quarry and stockpile volumetrics
• Construction earthworks tracking
• Agricultural field health mapping