Industries / Construction and property / Commercial and residential construction / Flooring
Carpet
Carpet installation is the work of laying broadloom (carpet made in wide rolls) or carpet tiles, stretching and fixing it to the floor, joining the seams, and taking the old carpet away.
7
Milestones achieved, from 1941 to 2023.
1
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1941
The tackless strip
Roy M. Roberts of Los Angeles patented a "carpet fastener" that held carpet without tacks: a narrow wooden strip about as thick as the carpet pad, nailed down on wood floors or glued on concrete, with rows of slanted prongs that hook the carpet's edge. Wall-to-wall carpet is still stretched over such strips.
Credit
Roy M. Roberts (US patent 2,238,946, filed April 1938, granted 22 April 1941)
c. 1950
Tufting makes carpet cheap
In Dalton, Georgia, makers of chenille bedspreads had adapted sewing machines for tufting, which pushes loops of yarn into a backing, and around 1950 Dalton firms widened these machines to cover a room-sized backing. In 1950 carpets were woven wool, at an average mill value of $6.24 per square yard; by 1970 tufted carpet made up 90% of the industry's volume and the average had fallen to $3.56. DuPont's bulked continuous filament nylon, introduced in 1957, performed close to wool at a lower price.
Credit
Dalton manufacturers and machine makers; DuPont. Figures from the New Georgia Encyclopedia
1973
Carpet comes in tiles
Ray Anderson founded Interface in 1973 to bring European carpet tiles to the US. Carpet tiles are squares of carpet laid one by one instead of from a roll, so a worn or stained one can be swapped. Between 1994 and 1999 Interface started ReEntry to take back its tiles for reuse and recycling, and today California's carpet stewardship program charges its fee on each square yard of both broadloom and carpet tile.
Credit
Ray C. Anderson and Interface
1987
The knee kicker's toll
A knee kicker is a short stretching tool the installer strikes with the knee. NIOSH, the US National Institute for Occupational Safety and Health, found that carpet layers were less than 0.06% of the US workforce but filed 6.2% of compensation claims for traumatic knee inflammation, and that frequent use of the kicker was the only significant predictor of bursitis, a swollen fluid sac at the knee. Measured peak kicks averaged 3,019 N in 1985, about four times body weight. The power stretcher, a mechanical alternative, gave about 14 times the kicker's mechanical advantage in a 1993 comparison, but a 1989 NIOSH survey found its absence from job sites was a major reason the kicker stayed in use. Danish floor layers with more than 30 years of kneeling work had higher odds of knee osteoarthritis on X-ray (odds ratio 4.82).
Lessona safer tool helps only if it is at the job.
pubmed.ncbi.nlm.nih.govpubmed.ncbi.nlm.nih.govpubmed.ncbi.nlm.nih.govpubmed.ncbi.nlm.nih.govpubmed.ncbi.nlm.nih.gov
Credit
Thun, Tanaka and colleagues at NIOSH; Bhattacharya and colleagues, University of Cincinnati; Village and colleagues, BC Research; L. K. Jensen and colleagues, Bispebjerg Hospital, Copenhagen
1992
Green Label after EPA complaints
Staff at EPA's own Washington headquarters, Waterside Mall, complained of ill health after a renovation that brought in new carpet. An EPA and NIOSH survey of 3,955 EPA workers (Wallace, Nelson and Dunteman, 1991) found that people in areas with new carpet reported more sore and dry throats and hoarseness. In 1992 the Carpet and Rug Institute (CRI), the US industry body, launched Green Label, which tests carpet, cushion and adhesives for emissions of volatile organic compounds (VOCs). In 1994 a District of Columbia jury awarded five EPA employees $948,000 against the building's owners, in a case that cited gases from new carpet containing 4-PC (4-phenylcyclohexene, an odorous chemical), paints and ventilation turned off during working hours. EPA's construction specification for its Research Triangle Park campus caps 4-PC in the air at 3 µg/m³ before the building is accepted.
Lessonbuyers now ask what a carpet gives off as well as how it wears.
ergo.human.cornell.educarpet-rug.orgenvironmentalhealth.caepa.gov
Credit
the EPA and NIOSH survey team; CRI
2010
A recycling fee per yard
California's AB 2398, signed on 30 September 2010, required carpet makers to run a recycling plan funded by a charge on each square yard sold in the state. The program is run by the Carpet America Recovery Effort (CARE), a nonprofit set up under an agreement signed on 8 January 2002 by the carpet industry, several states, EPA and a group of NGOs. CARE reported a recycling rate of 38.5% for 2024, collecting 82.7 million pounds of old carpet while 119.7 million pounds went to disposal, and in March 2026 CalRecycle, the state regulator, found that report noncompliant, noting that CARE has not been found fully compliant for any year of an approved plan. In 2022 Aquafil closed its new $12 million carpet recycling plant at Woodland, California, saying it could not source the 70 million pounds of used carpet a year needed to feed two plants.
Lessoncollecting old carpet, not building plants, has been the limit.
carpetrecovery.orgcarpetrecovery.orgwww2.calrecycle.ca.govresource-recycling.com
Credit
California Legislature; CARE; CalRecycle
2023
PFAS banned from carpet
PFAS, the "forever chemicals", have been used in stain and soil repellents for carpet. Maine banned the sale of carpets and rugs with intentionally added PFAS from 1 January 2023; New York's ban starts on 31 December 2026, and New Hampshire's and Rhode Island's on 1 January 2027.
Lessoncarpet lasts ten to twenty years, so treatments banned today will keep reaching recyclers for years (UC Berkeley, 2020).
legislature.maine.govdec.ny.govgc.nh.govdem.ri.govbcgc.berkeley.edu
Credit
the legislatures of Maine, New York, New Hampshire and Rhode Island
Watch for
Mixed nylon back to monomer
NylaNova, which licenses a Northwestern University process, says its catalyst turns post-consumer nylon-6 back into caprolactam, the raw material for new nylon-6, from mixed carpet, apparel and industrial scrap without pre-sorting, with yields above 90%. Its last tonne-scale pilot ran on feed that was only 65% nylon. It announced a supply deal with the nylon maker Nylene and plans a US plant making 20,000 tonnes a year from 2028, for which it is still raising money.
Achieved whenAn independent party confirms yield and purity on post-consumer carpet feed, and a commercial plant is running.
Claimed by
NylaNova, November 2026
Open
01
Count the kicks on jobs
A small logger fitted to a knee kicker's bumper that records every strike, with an accelerometer for timing and a thin force sensor for peak force, plus a short job log of room size and whether a power stretcher was on site. Installers do their normal work, and nothing about the work changes.
The briefCredit · unclaimed02
A quick PFAS screen
A test of whether cheap checks can tell PFAS-treated old carpet from untreated old carpet before it is recycled. Post-consumer samples go to a lab for combustion ion chromatography, which burns a sample and measures its fluorine, and the same samples are scored with an oil-drop repellency test, the manufacture date printed on the back, and the face-fibre type.
The briefCredit · unclaimed
Open 01 · the brief
Count the kicks on jobs
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A small logger fitted to a knee kicker's bumper that records every strike, with an accelerometer for timing and a thin force sensor for peak force, plus a short job log of room size and whether a power stretcher was on site. Installers do their normal work, and nothing about the work changes.
Why it is still open
Kick forces were measured with instrumented tools on 9 installers in 1985 and 7 in 1993, and the latest study of the tool, from 2012, tested bumper designs on a pendulum rig. Floor layers' kneeling is still scored by hand from video: one 2014 study analysed 45 videos of 32 floor layers. No source found gives a count of kicks per job.
Built with what exists
An Arduino Nano 33 BLE Sense or a similar board with a built-in accelerometer; a Tekscan FlexiForce thin-film force sensor; a pendulum rig like the one Huang and Wu built in 2012 for calibration; open-source peak detection in Python.
First step
Fit the logger to one kicker and check its count against 100 strikes on a pendulum rig.
Done looks like
A logger whose strike count is within 5% of a pendulum rig's count, then 20 logged installs, with a published table of kicks per job and per 10 m², split by whether a power stretcher was on site.
Open 02 · the brief
A quick PFAS screen
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A test of whether cheap checks can tell PFAS-treated old carpet from untreated old carpet before it is recycled. Post-consumer samples go to a lab for combustion ion chromatography, which burns a sample and measures its fluorine, and the same samples are scored with an oil-drop repellency test, the manufacture date printed on the back, and the face-fibre type.
Why it is still open
A 2020 UC Berkeley student report, written with advice from California's Department of Toxic Substances Control, found that PFAS were not removed, treated or monitored in carpet recycling, that the common screening method, particle-induced gamma-ray emission (PIGE), was not commercially available, and that recyclers should assume all old carpet is contaminated until there is a way to tell. A 2026 study still screened textiles with PIGE, which needs a particle accelerator.
Built with what exists
Commercial labs offering combustion ion chromatography (Centexbel is one); AATCC TM118 oil-repellency test liquids; a lab FTIR spectrometer to identify fibre type; scikit-learn for a simple classifier.
First step
Collect 20 samples from a CARE drop-off site and get quotes from two combustion ion chromatography labs.
Done looks like
100 post-consumer samples, each with a lab fluorine result and three quick-screen results, and the published sensitivity and specificity of each screen at a stated fluorine threshold.