SOLUTIONS · BY FACILITY

Materials recovery facilities: AI waste stream analysis, from infeed to residues.

Understand what comes in, what is recovered and what remains in the sorting residues. Waste stream analysis helps identify material losses, track variations and target the actions to check with the operator.

Challenges in 2026

Variable waste streams, costly residues, new risks.

Variable waste streams

Unpredictable incoming waste makes sorting harder and lowers material purity.

Residues and penalties

Sorting residues, up to 30% at some facilities, lead to lost recovery value and penalties.

Fire risk

Lithium batteries and hazardous contaminants drive up incidents and insurance costs.

Our solutions

Three responses, one set of field data.

Continuous waste stream analysis

On the line or off-line, real-time characterisation to identify material losses, track recovery and automate your reporting.

Contaminant detection

Identification of batteries, nitrous oxide canisters, metals and non-compliant plastics visible on the line, with an alert to the operator. They are still removed by hand, or by dedicated extraction equipment.

Fire risk management

Thermal patrol robots in storage areas and AI sensors on conveyors, to spot thermal anomalies and raise the alarm according to a scenario agreed with your teams.

Composition, live

See, continuously, what passes along your line.

Incoming waste, sorting residues or a recovered fraction observed on the conveyor before baling: depending on the measurement point chosen, the share of each visible category is estimated and recorded over time. When a critical visible contaminant appears, the operator is alerted.

  • Adjusting sorting settings using data tracked over time
  • Monitoring the quality observed before baling, useful for your customers and your audits
  • Traceability of contaminants and non-compliance
Residue lineContinuous

25%of recoverable materials identified in the residues

Breakdown of recoverable materials detected

  • Plastics 50%
  • Cardboard 30%
  • Aluminium 10%
  • Newspapers and magazines 10%
Contaminant detectedSmall electrical appliance · outgoing fraction
Example output. These values illustrate how the measurement works; they do not come from a client site.