{"stable_id":"5055c4918e7b18b5","slug":"reconfigurable-flow-chemistry","company":"MIT / University of Cambridge research team","workflow_name":"Reconfigurable self-optimizing continuous-flow chemistry","function_code":"engineering","pattern_codes":["autonomous_with_backstop"],"changed_assumption":"A machine-executable feedback loop can optimize reactions and produce a portable electronic protocol.","evidence_strength":"verified","publication_tier":"showcase","freshness":"current","reviewed_at":"2026-08-23","updated_at":"2026-08-22","source_quality_summary":"2 peer reviewed; publication outcomes are verified.","caveat_summary":"Research demonstration; no randomized comparison of labor or commercial production throughput.","freshness_caveat":null,"workflow_summary":{"before":[{"actor":"Synthetic chemist","action":"Selects reagents and repeatedly changes temperature, concentration, residence time, and other conditions.","actor_type":"control"},{"actor":"Chemist","action":"Interprets analytical results and chooses the next trial or reoptimizes a transferred procedure.","actor_type":"control"}],"hinge":"A machine-executable feedback loop can optimize reactions and produce a portable electronic protocol.","after":[{"actor":"Researcher and graphical interface","action":"Select reagents, unit operations, objectives, and constraints, then initiates optimization.","actor_type":"system"},{"actor":"Control and optimization software","action":"Runs reactors and analytics, selects subsequent conditions, and saves the optimized protocol.","actor_type":"system"}],"decision_mode":"bounded_autonomy","decision_marker":"AI acts within a human backstop"},"before":[{"order":1,"actor":"Synthetic chemist","action":"Selects reagents and repeatedly changes temperature, concentration, residence time, and other conditions.","handoff_to":"Analytical measurement","control":"Laboratory safety and chemist judgment; the paper describes this as labor-intensive trial-and-error."},{"order":2,"actor":"Chemist","action":"Interprets analytical results and chooses the next trial or reoptimizes a transferred procedure.","handoff_to":"Next experiment","control":"Manual iteration; no single manual cycle-time baseline is reported."}],"after":[{"order":1,"actor":"Researcher and graphical interface","action":"Select reagents, unit operations, objectives, and constraints, then initiates optimization.","handoff_to":"Automated flow platform","control":"Human-defined reaction and safety envelope."},{"order":2,"actor":"Control and optimization software","action":"Runs reactors and analytics, selects subsequent conditions, and saves the optimized protocol.","handoff_to":"Researcher or another apparatus","control":"Remote monitoring; electronic protocol transfer; out-of-bounds conditions require human intervention."}],"decision_rights":"Researchers choose chemistry, objective, and constraints; software owns bounded experimental sequencing.","exception_path":"Hardware faults, unsafe conditions, or unsupported chemistry stop optimization for researcher intervention.","removed_work":["Manual condition-by-condition iteration","Manual transcription of optimized conditions"],"outcomes":[{"metric":"Workflow breadth and optimization latency","baseline":"Labor-intensive trial-and-error and reoptimization","result":"Seven reaction classes plus a multistep sequence; optimization in hours or days","period":"Published 2018","scale":"More than 50 compounds in high yield across demonstrations","attribution_caveat":"Research demonstration; no randomized comparison of labor or commercial production throughput.","evidence_label":"verified"}],"executive_lesson":"The durable asset is not only the optimizer; it is an executable, transferable experimental protocol.","anti_pattern":"Do not equate laboratory breadth with autonomous invention of chemistry.","questions_for_leaders":["Where is the operating threshold set and who can override it?","What measured result would trigger rollback or retraining?","Which residual decisions must remain human-owned?"],"collections":[],"bundle_version":"1.0.0","bundle_fingerprint":"sha256:c23c6cc2b88153d008ea8fda928f632ce0011fc2d4c5036672a16e5d895bab93","canonical_url":"https://brianletort.ai/transformations/reconfigurable-flow-chemistry"}