The research library
284 peer reviewed works, chosen for one reason: each helps a buyer decide whether and how to automate a job. This is the research arm's foundation, and every entry links to the source so you can check us.
Most academic robotics is control theory, grasp planning and novel mechanisms, and none of it helps a forty person shop decide whether to automate a station. So we keep the other library: what a robot measurably does in a given application, what it costs and returns, whether automation costs jobs, whether smaller manufacturers see the same returns as large ones, and what happens to the operator.
By application
The job you are weighing, and what has been measured about it: throughput, accuracy against manual methods, reliability, cost and the people beside the machine. Our application guides and process pages cite these shelves directly.
- Palletizing and depalletizing 13 works
What the research measures about robots stacking and unstacking pallets: throughput, mixed loads, the manual lifting the robot replaces, and what it costs.
- Machine tending 10 works
Whether a robot loading CNC machines, presses and grinders actually runs unattended, how much part variety it tolerates, and what happens to the operator's job.
- Robotic and cobot welding 13 works
Weld quality and productivity against manual welding, sensitivity to fit-up, the welder shortage, and what welding cells cost and return.
- Pick and place, kitting and bin picking 14 works
How reliably vision-guided robots pick parts and items in real operations, where they fail, and how they compare with people.
- Packaging and end of line 13 works
What the research shows about automating case packing, bagging, labeling and the rest of the packaging line: line efficiency, changeover and the labor it replaces.
- Assembly, screwdriving and fastening 14 works
Measured results from robots and cobots doing assembly work: screwdriving, insertion, dispensing, how tasks are split with people, and the effect on cycle time and ergonomics.
- Machine vision and inspection on the floor 14 works
How automated visual inspection and metrology perform against human inspectors: false accepts and rejects, measurement accuracy, and NIST work on testing 3D vision.
- Drone inspection of structures 14 works
What drones measure on tanks, boilers, roofs, stockpiles and structures: accuracy against manual and survey methods, time, confined space exposure, and the limits.
- Material movement: AMRs, AGVs and tuggers 14 works
The evidence on mobile robots moving material in plants, warehouses and hospitals: fleet sizing, traffic, delivery performance and how staff work with them.
- Sanding, grinding, deburring and finishing 14 works
What robotic surface finishing achieves against manual finishing, and the dust, vibration and injury exposure of the manual job.
- Press brake and sheet metal handling 10 works
Research on robotic bending cells and sheet handling: cycle time, setup, part handling and when a bending cell pays.
- Security patrol robots and drones 11 works
What is known about robots and drones on patrol: deployments, detection, how people respond to them, and measured security outcomes.
- Floor cleaning robots 7 works
Evidence on robotic floor cleaning in commercial and public buildings: coverage, cleaning results, labor hours and how staff accept the machines.
- Robotic mowing and grounds care 14 works
What robotic mowers deliver on turf quality, energy, labor and safety, from turfgrass and engineering research.
- Service robots in hotels and restaurants 13 works
How guests and staff respond to service and delivery robots in hotels and restaurants, and what operators measured after deploying them.
By question
- Robots, jobs and wages 10 works
The evidence on what automation does to employment, which is the question every owner is really asking and almost nobody in this industry answers with a citation.
- Automation in smaller manufacturers 10 works
Adoption barriers, returns and failure modes specific to small and midsized firms, which behave nothing like the large plants most robotics research studies.
- Productivity, payback and what actually pays 10 works
Measured outcomes rather than vendor projections: productivity effects, investment appraisal, and the gap between the two.
- Line balancing, scheduling and the math Joe runs 10 works
The operations research underneath the concierge's quiet math: assembly line balancing, task allocation between people and robots, and cycle time under real constraints.
- Reliability, availability and OEE 10 works
Why a line of 98 percent machines is not a 98 percent line, and what the literature says about maintaining automated cells.
- Working alongside the machine 10 works
Human robot collaboration as it is actually experienced on a floor: safety, trust, workload and what happens to the operator's job.
- Skills, training and redeployment 10 works
What the evidence says about retraining the people already on the floor, which is the manifesto's central claim and one we should be held to.
- Drones for inspection: the evidence 10 works
What the literature measures about UAV inspection of bridges, tanks, sewers and structures: accuracy against manual methods, time, safety exposure, and where a drone still does not reach.
- Public safety technology, measured 10 works
License plate readers, gunshot detection, cameras and police drones as the evaluation literature actually finds them. The findings are mixed, and an agency deserves to see that before it buys.
- The economics above the atmosphere 10 works
Small satellites, launch costs, in-orbit servicing and the buy-the-data-or-fly-the-hardware question, from the literature rather than the pitch deck.
Metadata and abstracts come from OpenAlex under CC0, last harvested 2026-10-14, and every DOI was resolved in Crossref. We did not write the abstracts and we have not characterized anyone's findings. Joe can search this library mid conversation and cite what he finds, and he is instructed never to cite anything that is not in it.