Plastic sorting machines are the unsung heroes of modern recycling, transforming chaotic piles of waste into valuable resources. But just how many types of plastic can these technological workhorses distinguish? The answer depends on the technology, the complexity of the waste stream, and the precision required. In this guide, we’ll break down the common types of plastics sorted, the technologies that make separation possible, and the real-world capabilities of today’s sorting systems.
Understanding Plastic Resin Codes: The Foundation of Sorting
Before diving into sorting capabilities, it’s critical to recognize the 7 primary plastic resin types identified by the Society of the Plastics Industry (SPI) coding system. These codes, found on most plastic products, serve as the universal language for recyclers and sorting machines:
| Resin Code | Plastic Type | Common Uses | Recyclable? |
|---|---|---|---|
| 1 | Polyethylene Terephthalate (PET) | Soda bottles, food containers | Yes |
| 2 | High-Density Polyethylene (HDPE) | Milk jugs, detergent bottles | Yes |
| 3 | Polyvinyl Chloride (PVC) | Pipes, shower curtains | Limited |
| 4 | Low-Density Polyethylene (LDPE) | Grocery bags, squeeze bottles | Yes (with specialized systems) |
| 5 | Polypropylene (PP) | Yogurt cups, bottle caps | Yes |
| 6 | Polystyrene (PS) | Foam packaging, disposable cutlery | Limited |
| 7 | Other (e.g., PC, bioplastics) | 3D printing filaments, medical devices | Rarely |
Most basic sorting machines can target Resin Codes 1–5, while advanced systems handle 6 and select 7 types. The key challenge? Many plastics look identical to the human eye—think a PET water bottle vs. a PP medicine bottle—but have vastly different chemical properties.
How Plastic Sorters Identify Different Plastics
Modern sorting machines use a combination of spectral analysis, mechanical separation, and artificial intelligence to distinguish plastics. Here’s how the leading technologies work:
1. Near-Infrared (NIR) Spectroscopy
NIR is the workhorse of plastic sorting. It analyzes how plastics absorb and reflect near-infrared light, creating a unique “chemical fingerprint” for each resin. For example:
- PET absorbs light at 1.73 μm and 2.25 μm wavelengths.
- HDPE shows distinct peaks at 1.72 μm and 2.31 μm.
- PP has a signature absorption at 1.19 μm.
Capability: NIR systems can reliably separate 5–6 primary resin types (PET, HDPE, LDPE, PP, PS, and sometimes PVC) with 95%+ accuracy in clean waste streams. However, they struggle with dark or black plastics (which absorb NIR light) and multi-layered materials like chip bags.
2. AI-Powered Computer Vision
Newer systems pair high-resolution cameras with machine learning algorithms to analyze color, shape, and texture—even for black plastics. For example:
- Recycleye’s AI sorters use hyperspectral imaging to identify PP in black food packaging, achieving 92% purity in trials.
- TOMRA’s Autosort combines NIR with laser and electromagnetic sensors to sort 12+ plastic types, including difficult-to-detect composites.
Capability: AI-enhanced sorters can handle 8–10 distinct plastic types, including mixed-color and contaminated materials.
3. Electrostatic Separation
This technology uses triboelectric charging: plastics acquire unique charges when rubbed against materials like rubber or metal. For example:
- PVC and PET separate easily due to charge differences.
- HDPE and PP can be sorted with 90%+ purity in post-shredder processes.
Capability: Ideal for separating 2–3 similar plastics (e.g., HDPE vs. PP) after initial sorting by NIR or AI.
Real-World Sorting Capabilities: What Machines Actually Achieve
In practice, a plastic sorter’s performance depends on three factors: technology mix, waste stream complexity, and purity requirements. Here’s what to expect:
Basic Sorters (Municipal Recycling Facilities)
Most curbside recycling plants use NIR-only systems to separate 3–4 primary types:
- PET bottles (Code 1)
- HDPE containers (Code 2)
- PP caps (Code 5)
- Mixed plastics (Codes 4, 6, and 7 are often landfilled or incinerated).
Example: A typical U.S. MRF (Material Recovery Facility) sorts ~25 tons/hour, achieving 85–90% purity for PET and HDPE.
Advanced Sorters (Specialized Recycling Plants)
Facilities handling post-industrial or high-value plastics (e.g., food-grade recycling) use multi-technology systems to separate 6–8 types:
- PET flakes (for bottle-to-bottle recycling)
- HDPE and PP (for injection molding)
- PS foam (for insulation)
- Even niche plastics like nylon (PA) and polycarbonate (PC).
Case Study: TotalEnergies’ Project Omni uses AI-powered sorting to separate food-grade PP from household waste with 95% purity, enabling its reuse in new food packaging.
Cutting-Edge Research Systems
Laboratories and pilot projects are pushing boundaries:
- The University of California, Berkeley, developed a sorter using Raman spectroscopy to identify 12+ plastic types, including bioplastics like PLA.
- Dutch firm BrightLoop uses digital watermarks embedded in packaging to help sorters identify even multi-layered plastics.
Challenges Limiting Sorting Capabilities
Despite advances, several hurdles prevent sorters from separating all plastics:
1. Contamination
Food residue, labels, or non-plastic materials (e.g., metal caps) can confuse sensors. The EPA estimates 25% of recycling bins are contaminated, reducing sorting accuracy by 10–15%.
2. Black Plastics
Carbon black pigment absorbs NIR light, making traditional systems blind to black PP or PS. AI and laser-based systems now solve this but add 20–30% to equipment costs.
3. Multi-Layered Plastics
Packaging like chip bags (combinations of PET, aluminum, and LDPE) is nearly impossible to sort into pure streams, even for advanced machines.
4. Cost vs. Value
Sorting low-value plastics (e.g., PVC, Code 3) often costs more than the recycled material is worth, limiting commercial adoption.
Future Trends: More Types, Higher Purity
Innovation is expanding sorting capabilities:
- AI and Machine Learning: Systems like Recycleye’s can now learn to identify new plastic types with minimal human input, potentially handling 15+ resins by 2030.
- Hyperspectral Imaging: Cameras capturing 100+ wavelengths will distinguish even subtle differences (e.g., recycled PET vs. virgin PET).
- Chemical Tracers: Companies like Nextek are developing invisible “barcodes” for plastics, allowing sorters to identify specific brands or formulations.
Conclusion: How Many Types Can Be Sorted?
Today’s best plastic sorters can reliably separate 8–10 distinct plastic types under ideal conditions. For most recycling facilities, the practical number is 5–6 common resins (PET, HDPE, LDPE, PP, PS, and PVC). As AI and sensor technologies advance, this number will grow—bringing us closer to a truly circular plastic economy.
The next time you toss a plastic bottle into the recycling bin, remember: behind the scenes, a symphony of technology is working to give it a second life. And with each innovation, we’re one step closer to sorting all plastics, not just the easy ones.
Comments(18)
Wow, had no idea sorting plastics was this complex! The NIR spectroscopy part blew my mind 🤯
This explains why my recycling bin sometimes gets rejected – probably contaminated with the wrong plastics.
Fascinating read! The part about AI identifying black plastics is game-changing for recycling centers.
Anyone else surprised that PVC is so hard to recycle? I thought pipes would be easier to process.
@BlackthornMystic：Right? PVC contains chlorine which makes it toxic when melted down. That’s why most recyclers avoid it.
Great breakdown! Now I understand why those “recyclable” coffee cups aren’t actually getting recycled.
The future tech with chemical tracers sounds promising. Maybe we’ll finally solve the multi-layer packaging issue!
@Roaring Storm：Those chemical tracers seem like the holy grail! Imagine if every plastic had an invisible ID tag – sorting would be so much easier.
As someone who works at a recycling facility, this article nails the challenges we face daily with plastic sorting.
@CelestialFox：As a fellow facility worker, the contamination stats hit home. People don’t realize how much their pizza boxes ruin entire batches 😅
I wish my local recycling center had those advanced AI sorters. Our town only takes #1 and #2 plastics 😕
The section on electrostatic separation was super technical but really interesting. Science is cool!
This makes me want to be more careful about separating my plastics properly before recycling.
The breakdown of different sorting technologies is super helpful! Didn’t know NIR was that accurate for common plastics 👍
This makes me wonder – why don’t more facilities invest in those AI sorters if they can handle black plastics? The cost factor mentioned explains a lot.
The part about multi-layered plastics being nearly impossible to recycle explains why my local center stopped taking chip bags. So frustrating!
Fascinating tech, but we should really focus on reducing plastic use first. No sorting machine can keep up with our current waste levels 🤷♂️
That Berkeley research sounds promising! Hope these innovations trickle down to local recycling centers soon.