Selecting a collaborative robot is a multi-criteria decision where payload, reach, repeatability, number of axes, and safety-rated monitoring dominate the shortlist, with quoted prices running from roughly $10,000 for a basic 4-axis arm to over $100,000 for a >20 kg payload class with vision and torque sensing [S3].
The category has been structured into a formal MCDM (multi-criteria decision-making) framework using the Analytical Hierarchy Process, treating cobot selection as a weighted score across technical and economic criteria rather than a single-vendor comparison [S2]. Pricing tiers, growth rates, and use-case patterns from UFACTORY and IDTechEx put 2025 market value near $2.14B and a CAGR of 31.6% to 2030, with 3–5% annual price erosion expected as competition intensifies [S3][S5].
What a Cobot Actually Is, and Where the Boundary Sits
A cobot is a slow-moving, lightweight manipulator designed to share workspace with a human operator without a physical fence, a definition that the IDTechEx 2025–2045 market study uses to scope the entire product class [S5]. The defining engineering claim is power-and-force-limiting (PFL) at the joint or surface, plus a safety-rated monitored stop, which together let the unit comply with collaborative-mode requirements in ISO 10218-1 and the technical specification ISO/TS 15066 [S2].
That boundary is also where buyers get burned. A standard industrial 6-axis articulated robot running at full speed with the guard removed is not a cobot, no matter what the marketing says, because the safety architecture (category-3 STO, monitored stop, PFL on every axis) is what makes the cell collaborative. Spec sheets from mainstream vendors — ABB's GoFa and SWIFTI families, for example — explicitly separate the collaborative line from the fenced industrial line on this basis [S1].
The Four Decision Criteria That Drive the Shortlist
Payload capacity is the single largest price multiplier: cobots rated above 20 kg cost more than twice as much as sub-5 kg units, per ABI Research data republished in the UFACTORY pricing study [S3]. For a 3 kg pick-and-place cell on a packaging line, a 12 kg SCARA-class cobot is overkill and wastes both budget and footprint; for a 25 kg palletizing duty, a 5 kg tabletop arm is unsafe regardless of cycle-rate claims.
Reach is the second axis. A 500 mm working envelope is right for benchtop electronics assembly; a 1300–1400 mm reach is the minimum to cover a half-pallet or a machine-tending cell without rail mounting. Repeatability sits in the ±0.02–0.05 mm band for most 6-axis cobots, with sub-±0.02 mm reserved for high-precision screw-driving and dispensing cells; buyers should not pay for precision they cannot measure on the shop floor [S3].
Number of axes is the third. Standard 6-axis arms dominate the market because they cover the largest task envelope; 4-axis SCARA-type units drop to roughly $8,000 and suit simple top-down pick-and-place, while 7-axis units exist for confined-space welding and inspection but carry a price premium driven by the extra joint and cable management [S3]. For a side-by-side articulated-class comparison, our SCARA robot spec map walks through the reach/repeatability trade-off in adjacent detail.
When a Cobot Is the Wrong Tool
High-throughput palletizing at cycle times under 6 seconds, heavy-part welding, and cleanroom-class electronics pick-and-place with sub-second indexing are all cases where a dedicated industrial articulated robot or a high-speed [delta robot](/encyclopedia/delta-robot.html) beats a cobot on cost-per-part. IDTechEx flags the cobot's deliberate speed limitation as a structural ceiling on duty cycle, which is why a growing share of new cells pair a cobot with a wheeled AMR for part delivery, splitting the slow/fast tasks across two platforms [S5].
Cobots also do not suit heavy-payload (>35 kg) material handling. The Proxie launch from Collaborative Robotics in November 2024 explicitly targets the 20–40 kg class of material-handling tasks, filling the gap between tabletop cobots and traditional industrial arms, and is indicative of where the market is segmenting between cobot and AGV-class mobile platforms [S4].
Total Cost of Ownership, Not Sticker Price
The purchase price covers roughly half the five-year cost on most cobot deployments. Hidden-cost line items that flip the shortlist are: end-effector and gripper integration, safety scanners and area-muting light curtains if the cell needs collaborative/traditional hybrid mode, fixturing, and a 1–2 week integration effort that vendors price separately or bundle depending on channel [S3].
Service contracts, spare joint modules, and the cost of a 1–2 day operator training class per shift are recurring TCO drivers that buyers frequently underestimate. A realistic TCO envelope for a 6-axis collaborative cell sits 1.4–1.8× the unit price across a 60-month lifecycle, with the multiplier higher for vision-enabled and force-torque-enabled configurations [S3].
Application Mapping by Use Case
Machine tending of CNC lathes and presses is the largest single cobot application, dominated by 6-axis arms in the 5–12 kg payload class with reach between 900 and 1300 mm. For assembly of consumer electronics, sub-±0.03 mm repeatability and a force-torque sensor at the wrist are non-negotiable; this is where the 7-axis and high-precision 6-axis units earn their price premium [S3][S5].
Welding cobots are a distinct sub-segment with torque sensing and a through-arm dress package for MIG torches. Palletizing cobots cluster at the 12–30 kg payload band, often paired with a seventh axis linear rail to extend reach; for higher throughput, an industrial articulated robot on a fixed pedestal will out-cycle a cobot by a factor of 2–3. For adjacent coverage of spec-led selection logic on packaging lines, our delta robot spec map is a useful cross-reference on high-speed pick-and-place.
Selection Logic, in Five Lines
Step 1: lock payload and reach from the worst-case part, not the average part — the >20 kg class doubles the unit price [S3]. Step 2: decide between 4-axis SCARA, 6-axis, or 7-axis from the task envelope; a 4-axis unit is fine for top-down assembly but cannot reach around obstacles. Step 3: confirm collaborative-mode certification and PFL on every axis against ISO/TS 15066 and ISO 10218-1, not just a vendor white paper [S2][S5]. Step 4: budget TCO at 1.4–1.8× unit price across five years, including integration, training, and end-effector tooling [S3]. Step 5: shortlist two or three vendors and run a paid pilot on the actual part before signing the PO; the >$100,000 high-end segment is where vendor performance diverges most [S3][S4].
For buyers still framing the question, IDTechEx's 2025–2045 forecast and the UFACTORY 2025 pricing study remain the two most-cited public reference documents in the segment [S3][S5]. Track the 2025 unit-shipment data from the IFR World Robotics report and the next round of OEM safety certifications for ISO/TS 15066 conformance as the next decision-shaping data points.