A Tier-Graded Alternative to Binary Pass/Fail Marks for Cannabis Consumption Hardware
Author: Matt Macosko, Founder, Cannabis Device Safety Institute Affiliation: Cannabis Device Safety Institute, Arcata, California Status: Founding draft, v0.4 (May 2026) Suggested citation: Macosko, M. Provenance Certification: A Tier-Graded Alternative to Binary Pass/Fail Marks for Cannabis Consumption Hardware. CDSI Working Paper P-006. 2026. Companion documents: P-001 The Hardware Vacuum; P-002 Loaded-State Off-Gas Analysis; P-003 The Pyrolytic Reservoir.
Status of the certification mark — read before citing this paper. This paper is a design proposal, not a description of an operating program. The CDSI Provenance Certification mark does not currently exist. No device has been certified by CDSI at any tier, and no CDSI registry is operating. The Institute does not yet have a testing bench, and has issued no certification of any kind to any manufacturer — including hardware built by the founder’s own companies. The tiers, obligations, and governance mechanisms described below are what CDSI proposes to build and hold itself to; every use of “Tier 1/2/3” in this document is prospective. Nothing here should be read as a claim that any product carries any CDSI mark.
Abstract
Most safety certification marks in consumer-facing hardware markets are binary: a product either carries the mark or it does not. Binary marks are administratively cheap and visually clear, but they collapse meaningful gradations of safety information into a single bit. For an emerging field like cannabis consumption hardware — where the underlying analytical methodology is itself in active development, where manufacturer testing capacity varies by an order of magnitude across the industry, and where consumer harm-reduction value scales with the depth of characterization — a binary mark is the wrong instrument. This paper argues for a tier-graded provenance certification model that publishes what was tested, by whom, and how, rather than collapsing those facts into a single seal. We specify the three-tier model proposed by the Cannabis Device Safety Institute — Tier 1 Reported, Tier 2 Certified, Tier 3 Gold Standard — defining the obligations, allowable claims, and consumer-facing presentation at each tier. We address governance for tier promotion and demotion, dispute resolution, and the conflict-of-interest mechanisms that prevent tiering from collapsing into industry capture. We close with the economic case for tiering: it admits resource-constrained manufacturers into the system as Tier 1 participants while preserving Tier 3 as an aspirational standard whose certification cost is justified by its informational density. The model draws on mature precedent from the U.S. ENERGY STAR program, NSF International’s certification levels, and the European olive oil grade hierarchy. It does not require legislative action to deploy.
Keywords: certification design, tiered certification, hardware safety standards, consumer information disclosure, regulatory institutional design, voluntary certification
1. Introduction
The Cannabis Device Safety Institute’s working position is that the cannabis consumption hardware safety field is not a missing-regulation problem; it is a missing-institution problem (P-001). Working paper P-002 specified the analytical methodology — loaded-state off-gas analysis — that we propose as the new minimum for hardware characterization. Working paper P-003 extended that methodology to address cycle-life effects, particularly the pyrolytic reservoir.
The present paper concerns the certification mechanism by which the results of those tests reach consumers, regulators, retailers, and journalists. The claim of this paper is sharper than it may first appear: even if every device on the market were correctly tested under CDSI-001, the field would still be poorly served by the dominant certification idiom in adjacent industries — the binary pass/fail mark.
A binary mark presents only one bit of information. The mark is present, or it is absent. A device that has been characterized in dry-fire conditions at one specific power level by an in-house QA team carries the same visual signal as a device that has been characterized across three matrices, four cycle-count regimes, and an interlaboratory replication study. The two are not equivalent claims, and a certification system that treats them as equivalent does affirmative damage to the field by suppressing distinctions that consumers, regulators, and engineers all need.
This paper argues for a deliberate alternative: a tier-graded certification model in which the visible mark always indicates which level of certification the device has earned, and the public registry behind the mark documents what was tested, by whom, and under what protocol. We call this approach provenance certification, by analogy to provenance disclosure in art markets and food origin labeling. Our claim is not that provenance certification is a CDSI invention; it is not. Our claim is that it is the right design pattern for a field whose underlying science is still developing and whose participant capacity is heterogeneous.
The paper proceeds as follows. Section 2 catalogs the specific failure modes of binary certification when applied to cannabis hardware. Section 3 reviews tier-graded certification in three adjacent fields and extracts design principles. Section 4 specifies the CDSI three-tier model formally, including the obligations and allowable claims at each tier. Section 5 elaborates the obligations of each tier in operational detail. Section 6 addresses the governance question that tiering raises but does not answer: who decides which tier a product earns, and how is tier movement adjudicated. Section 7 discusses the economics of tier pricing and the deliberate cross-subsidy structure that supports the model. Section 8 catalogs known limitations and the design questions that should be revisited as the model is exercised.
2. Why binary marks fail consumers and regulators in this field
2.1 The underspecified mark
The most pervasive failure mode of binary safety marks is consumer ambiguity about what the mark actually attests to. A product carrying a UL mark, for example, is certified for one specific scope — typically electrical safety against fire and shock. A consumer is not generally aware that the same UL mark says nothing about the chemical composition of the product’s emissions, the durability of its sensors, the manufacturing-batch consistency of its outputs, or the cybersecurity of its embedded firmware. The mark’s visual simplicity is, in effect, a misrepresentation of its scope.
In the cannabis hardware case, the consumer is in a structurally worse position. The relevant scope of “safety” is not yet stable enough for a single mark to be unambiguous. A device that has been tested for electrical safety, materials composition, dry-fire emissions, loaded-state emissions on a single matrix, loaded-state emissions across multiple matrices, and cycle-life characterization is plausibly safer-known than a device that has been tested only against the first two. A binary mark cannot communicate that gradient.
2.2 The all-or-nothing accession problem
Binary marks force a manufacturer into a discrete choice: meet the full bar or be uncertified. For an industry whose participants range from the venture-funded multi-million-unit incumbents to the single-employee niche makers serving harm-reduction communities, the bar set by a binary mark will either be too low (admitting low-rigor manufacturers who undermine the mark’s value) or too high (excluding small manufacturers whose products may actually be safer than some certified incumbents but who lack the capital to produce the certification dossier). Neither outcome is desirable.
The accession problem also creates an incentive distortion. A manufacturer who has invested substantially in characterization but cannot quite reach the binary mark’s bar receives no recognition for the work they have done. The mark therefore fails to incentivize incremental progress; it incentivizes only the discrete jump from uncertified to certified.
2.3 The compressed-signal problem for regulators
State regulators considering whether to require a particular safety mark in their statutes are, in effect, being asked to endorse a discrete cutoff: any product with the mark is acceptable, any product without it is not. This binary framing collapses a regulatory decision that should reflect risk-graded thinking — high-risk products held to the highest tier, low-risk products held to a lower tier — into an artifact of the certification system’s design. Regulators with the resources to draft tier-graded statutes (California, New York, Colorado) end up doing so themselves, because the certifications they would otherwise reference are too coarse-grained to map onto risk-stratified policy.
2.4 The compressed-signal problem for consumers
The argument is symmetrical for consumers. A consumer comparing two products at retail receives the same visual signal from a Tier-3 Gold-Standard-equivalent product and a Tier-1 Reported-equivalent product if the certification system collapses both to a single mark. The certification system has, in this case, deprived the consumer of information that the underlying testing actually generated.
2.5 The consequence: certification is not what it claims to be
The cumulative effect of these failure modes is that binary certification marks, in fields where the underlying science is heterogeneous and the participant base is varied, underperform their stated function. They claim to inform; they actually compress. They claim to incentivize safety improvement; they actually incentivize a single discrete jump. They claim to support regulatory mapping; they actually frustrate it. The remedy is to design a certification system whose visual mark and public registry are both tier-graded.
3. Adjacent-field analogues
We are not the first to encounter these problems, and the design pattern we propose has substantial precedent in adjacent fields. Three are particularly instructive.
3.1 ENERGY STAR and “Most Efficient”
The U.S. Environmental Protection Agency’s ENERGY STAR program is a binary mark — a product either qualifies as ENERGY STAR or it does not — but it has been progressively augmented since 2011 with a tier above it: ENERGY STAR Most Efficient, awarded annually to the top performers in each category. The two marks coexist on the same products, and the consumer-facing presentation distinguishes them visually. This bolted-on second tier was a direct EPA response to manufacturer feedback that the binary ENERGY STAR mark had ceased to differentiate at the high end of the market.
The lesson for our purposes is that the EPA discovered, the hard way and over a multi-year retrofit, that a binary mark in a maturing market produces an undifferentiated upper tier. ENERGY STAR Most Efficient is the EPA’s answer to that problem. CDSI, designing from a clean slate, can incorporate the lesson at the front end.
3.2 NSF International’s tiered certification levels
NSF International, the standards body that certifies food-equipment safety in the U.S. and abroad, has used a multi-tier certification model for decades. A product can be certified to NSF/ANSI 51 (food-contact materials), to NSF/ANSI 4 (commercial cooking equipment), or to NSF Sanitation Compliance — each with distinct scope, methodology, and consumer-facing presentation. The tiers are not vertically ordered; they are domain-specific. But the underlying design is the one we draw on: a single institutional source can issue multiple distinct certifications whose differences are publicly documented.
The lesson here is that a single certifying body can credibly maintain multiple distinct marks if the marks’ meanings are documented in the public registry and the body’s own governance prevents tier conflation.
3.3 European olive oil grade hierarchy
The European Union’s olive oil regulation (EC 2568/91 and successors) specifies a tier hierarchy: extra virgin > virgin > refined olive oil > olive pomace oil. Each tier carries distinct labeling requirements, allowable processing methods, and analytical-chemistry pass criteria (free fatty acid content, peroxide value, sensory panel scores). The labels are mandatory, the chemistry is published, and the consumer can navigate the hierarchy on the bottle.
The European model is in some ways the closest analogue to what we propose, because olive oil — like cannabis hardware — has both an analytical-chemistry component (chemical specifications) and a process component (how the oil was produced). CDSI’s tiers similarly span both dimensions: a Tier 3 device has both met higher analytical-chemistry bars and undergone a more demanding process (production-lot re-testing, registry publication, etc.). The olive oil precedent demonstrates that consumers can absorb a four-tier hierarchy if the labels are consistent and the underlying tests are public.
4. The CDSI three-tier model
The model proposed is named CDSI Provenance Certification. It comprises three tiers. Each tier carries a distinct visual mark, a defined scope of what the mark attests, and a corresponding entry in the public CDSI registry.
4.1 Tier 1 — Reported
Definition: The manufacturer has performed the CDSI baseline analytical schedule (CDSI-001 §1–6) in-house or via a non-accredited laboratory and has submitted the results to the CDSI registry. CDSI has not independently verified the results.
Visual mark: Single-line “CDSI Reported” lockup, no badge ornamentation. The visual treatment is intentionally austere to communicate that the mark is a self-reported claim, not an accredited verification.
Allowable claims: The manufacturer may state that the device has been “Reported under CDSI baseline methodology” and may reproduce its registry entry on packaging and marketing materials. The manufacturer may not state that the device is “CDSI Certified” or “CDSI Gold Standard.”
Registry entry: Includes the test report, the testing facility’s identity and qualifications, the lot or unit identification of the tested device, the date of testing, and any deviations from the CDSI-001 protocol.
Re-testing cadence: None required. Tier 1 entries persist in the registry indefinitely with their original test date prominently displayed.
4.2 Tier 2 — Certified
Definition: A CDSI-accredited third-party laboratory has performed the CDSI baseline analytical schedule (CDSI-001 §1–6) on the device, and the device’s results meet the publicly documented Tier 2 acceptance criteria (which are themselves a separate CDSI document subject to revision).
Visual mark: Full “CDSI Certified” badge with version year. The badge is designed to be visually distinct from Tier 1 and from Tier 3.
Allowable claims: The manufacturer may state that the device is “CDSI Certified (year)” and may reproduce its registry entry. The manufacturer may not state any claim that conflates Tier 2 with Tier 3 or implies cycle-life characterization that has not been performed.
Registry entry: Includes the accredited lab’s identity and accreditation reference, the full test report, the device’s lot or product-line identification, and any deviations from the CDSI-001 protocol.
Re-testing cadence: Every two years from the certification date, or upon any material revision to the device’s design or supply chain. CDSI publishes a notice when a Tier 2 certification expires; a manufacturer who does not re-test loses the Tier 2 mark.
4.3 Tier 3 — Gold Standard / Ongoing Batch
Definition: Tier 2 plus (a) cycle-life characterization per the proposed CDSI-001 §7 (P-003), and (b) periodic re-testing of production lots from the manufacturer’s actual output, with results published to the registry, and (c) public registry of cycle-life data accessible to consumers, regulators, and harm-reduction researchers.
Visual mark: Full “CDSI Gold Standard” badge with version year and a small “Ongoing Batch” indicator. The badge is the visually most prominent of the three, reflecting both its informational density and its operational cost.
Allowable claims: The manufacturer may state that the device is “CDSI Gold Standard / Ongoing Batch (year)” and may reproduce both the certification dossier and the cycle-life summary on packaging and marketing materials.
Registry entry: Includes everything in the Tier 2 entry, plus the cycle-life data, plus the published production-lot test results to date, with the lot-testing cadence and methodology documented.
Re-testing cadence: Quarterly production-lot re-testing, plus a full recharacterization every two years.
4.4 What the tiers are not
Three clarifications:
- The tiers are additive, not graded scores within the same scope. A Tier 3 device is not “a Tier 2 device that scored higher”; it is a device that has been characterized over a broader scope. The distinction matters for consumer presentation: a Tier 3 mark says “we know more about this device,” not “this device is safer than a Tier 2 device.”
- The tiers do not exempt manufacturers from any other certification regime that may apply to their product. A Tier 3 cannabis vaporizer must still comply with applicable state regulations, federal regulations where they apply, and any other certification marks the manufacturer may also wish to carry (UL 8139, etc.).
- The tiers are device-level, not manufacturer-level. A manufacturer may have devices at all three tiers in their product line, and the registry is queryable at the device-SKU level rather than at the company level.
5. Tier obligations in operational detail
This section elaborates the obligations attached to each tier, in language sufficient for an operationalizing protocol document. The full operational protocol is in development and will be published as CDSI-002 alongside its first revision.
5.1 Manufacturer obligations
At all tiers, the manufacturer must: (a) provide the test laboratory with production-stock units (not specially fabricated samples), (b) document and disclose any modifications to the device in the period between testing and certification publication, (c) retain documentation supporting the test for the duration of the certification’s validity plus seven years, and (d) cooperate with any CDSI-initiated audit of the underlying test records.
At Tier 2 and Tier 3, additional obligations apply: the manufacturer must notify CDSI of any material design change, supply-chain change, or contract-manufacturing change within thirty days; failure to do so is grounds for tier demotion.
At Tier 3, the additional obligation is the production-lot re-testing schedule: the manufacturer must submit production-stock units from each calendar quarter to a CDSI-accredited laboratory for re-testing against the Tier 3 schedule, and the results are published in the registry within ninety days of testing.
5.2 Laboratory obligations
For Tier 2 and Tier 3, the testing laboratory must hold CDSI accreditation. CDSI accreditation comprises (a) demonstrated competence with the CDSI-001 protocol via a proficiency-testing program, (b) documented quality-management practices, (c) annual on-site or virtual audit by CDSI, and (d) freedom from conflict of interest with the certified manufacturer (no fee structure that ties laboratory revenue to certification outcome, no equity or affiliate relationship).
The accreditation body is CDSI itself for the early years of operation. As the field matures, CDSI intends to delegate the accreditation function to an independent accreditation body (the same delegation pattern by which national accreditation bodies like ANAB hold individual labs accountable to ISO 17025).
5.3 CDSI obligations
CDSI commits to: (a) maintaining the public registry in a queryable, machine-readable format with published API access; (b) publishing methodology documents, acceptance criteria, and revision history under an open license that permits reproduction and translation; (c) handling tier-promotion and tier-demotion requests within a published timeline; (d) publishing an annual report on certification activity, including aggregate statistics on tier distribution by device class.
CDSI will not certify products in which CDSI’s founder, board members, or staff have a material financial interest. The conflict-of-interest mechanism is documented in §6.
6. Governance: tier promotion, demotion, and dispute resolution
6.1 Promotion
A device’s tier may be promoted (Tier 1 → Tier 2, Tier 2 → Tier 3) at any time upon the manufacturer’s submission of the additional testing required at the higher tier. Promotion is administrative; CDSI verifies that the higher-tier obligations have been met and updates the registry. The visual mark is updated by the manufacturer at their next packaging revision.
6.2 Demotion
A device may be demoted by CDSI under three conditions:
- Failure to maintain re-testing cadence. Tier 2 lapses to “expired Tier 2” if the two-year re-test is not submitted by its due date. Tier 3 lapses similarly, plus quarterly production-lot test failures cause Tier 3 to demote to Tier 2.
- Material undisclosed change. Discovery (whether by CDSI audit, manufacturer self-report, or third-party reporting) of a material design or supply-chain change that was not disclosed within thirty days demotes the device by one tier and is published in the registry.
- Audit failure. A CDSI audit that identifies materially incorrect test records or noncompliance with the protocol demotes the device. The demotion is published with the audit findings.
Demotion is appealable by the manufacturer through a published appeals procedure (CDSI Standard Operating Procedure 06, in development).
6.3 Dispute resolution
Disputes over tier classification, accreditation status, audit findings, or registry accuracy are handled by a CDSI Standards Review Panel. The panel comprises three members: one CDSI staff member, one external standards-body representative (ASTM, NSF, UL, or equivalent), and one consumer or harm-reduction-community representative. Panel decisions are public. The panel does not adjudicate scientific disputes about the underlying methodology; those go through the methodology revision process.
6.4 Conflict of interest
CDSI’s Conflict of Interest Policy treats Founder, director, officer, and staff financial interests under a disclosure-plus-structured-recusal regime rather than as outright bars to participation. The Founder Director’s continuing ownership and operational roles in Divine Tribe and Nice Dreamz LLC are acknowledged on the record at the time the Policy was adopted (COI Policy, Article II, Section 2.4) and require no per-meeting re-disclosure; any CDSI or LLC Subsidiary action specifically affecting Divine Tribe or Nice Dreamz LLC products, certifications, or contracts triggers Founder recusal under Article III, Section 3.4 (sole-director special procedure: independent review by outside counsel or a disinterested advisor, fair-market-value pricing, written record in the corporate minute book, and ratification by the full Board after additional directors are seated). The same disclosure-plus-recusal regime applies prospectively to any future director, officer, or staff member with a Hardware-Industry Interest. The cost of institutional credibility is paid by the procedure, not by exclusion.
The accreditation function — qualifying laboratories — is held to the same conflict-of-interest standard. CDSI does not accredit laboratories owned by, or financially dependent on, CDSI staff or board.
7. Economics: tier pricing and cross-subsidy
The CDSI tier system is designed to be self-sustaining within five to seven years of operation and is structured around a deliberate cross-subsidy.
7.1 Pricing structure (proposed)
- Tier 1 — Reported: Registry submission is free for the first two years of CDSI operation; thereafter a modest registry-maintenance fee covers the cost of the registry infrastructure. CDSI does not perform analysis at Tier 1 and incurs no laboratory cost.
- Tier 2 — Certified: The manufacturer pays the accredited laboratory directly for the analytical work (typical lab cost on the order of $5,000–$15,000 per device per characterization cycle). CDSI charges a per-certification administrative fee (currently estimated at $1,500 per device per two-year certification cycle), which covers registry publication, audit overhead, and accreditation maintenance.
- Tier 3 — Gold Standard / Ongoing Batch: Tier 2 costs apply, plus the Tier 3 cycle-life characterization (estimated lab cost $15,000–$30,000 per device per recharacterization cycle), plus the quarterly production-lot re-testing (estimated lab cost $2,000–$5,000 per quarter per SKU). CDSI’s administrative fee at Tier 3 is currently estimated at $5,000 per device per two-year cycle.
7.2 Cross-subsidy structure
Tier 3 fees are set above CDSI’s marginal cost; Tier 1 fees are set at or near zero. The intent is that the larger manufacturers, who can absorb Tier 3 cost as a marketing investment and benefit most from the Tier 3 mark’s premium positioning, partially fund the registry infrastructure that admits smaller manufacturers at Tier 1 with no financial barrier.
This cross-subsidy is a deliberate institutional choice. The alternative — pricing Tier 1 to recover its own administrative cost — would price out the harm-reduction makers and small-batch manufacturers whose participation makes the registry useful as a comprehensive industry view rather than a yacht club for the largest players.
7.3 Why CDSI does not perform Tier 2 or Tier 3 testing in-house
CDSI is designed as a certifying body, not as a contract laboratory. The accredited laboratories that perform the analytical work are independent commercial entities. This separation is a structural conflict-of-interest mitigation: CDSI’s institutional incentive should be the credibility of the certification, not the volume of testing it performs. A certifying body that performs its own paid analytical work has a structural conflict that no governance procedure fully resolves.
CDSI may, in early years, operate a reference laboratory for proficiency testing and methodology development. The reference laboratory does not certify commercial products.
8. Limitations
8.1 Tier 1 risks misuse as a credential laundering mechanism
The single largest risk of including Tier 1 in the model is that manufacturers may use “CDSI Reported” status as a marketing credential without communicating to consumers that no independent verification has occurred. The mitigations specified in §4.1 (austere visual mark, mandatory registry entry containing testing-facility identity and qualifications) reduce this risk but do not eliminate it. The risk should be monitored, and the model should be revisited if Tier 1 status is observed in practice to be marketed in misleading ways.
8.2 Visual differentiation is harder than it sounds
The proposal that the three tier marks be “visually distinct” is operationally non-trivial. Mark design must balance (a) instant consumer recognition, (b) clear hierarchy, (c) accessibility for color-blind and low-vision consumers, and (d) reproducibility on small package surfaces (cartridge boxes, atomizer sleeves, etc.). CDSI’s first design draft is in development and will be subjected to external review before publication.
8.3 The two-year Tier 2 re-test cadence may be wrong
The choice of a two-year re-test interval at Tier 2 is informed by adjacent-field practice (NSF certification re-test cadence) but not derived from a model of how cannabis hardware’s analytical characterization should be expected to drift over time. Empirical data from the first cohort of Tier 2 certifications will inform whether the cadence should tighten or loosen.
8.4 The cross-subsidy may not be sustainable in the published-fee structure
The fee estimates in §7.1 are first-pass and assume operational scale that CDSI has not yet achieved. The cross-subsidy may need to be supplemented in early years by foundation funding or grant support. CDSI’s status as a 501(c)(3) public charity (IRS determination letter dated 2026-06-29; § 509(a)(2); exemption effective 2026-04-27) is the institutional vehicle for that supplementation.
8.5 The model assumes industry voluntary participation
CDSI tiering is voluntary and market-incentivized. The model does not address the case in which a state or federal regulator wishes to require CDSI certification at a specified tier. We believe a regulator-required tier could be accommodated within the existing model — the regulator simply specifies that, for example, all products sold in their jurisdiction must hold at least Tier 2 status — but the operational implications of becoming a regulator-mandated certification body are substantial and should be considered by CDSI’s board if and when that conversation arises.
8.6 The model does not cover services, only products
CDSI Provenance Certification, as specified, applies to physical devices. Service offerings (vape lounges, dispensary on-site product use, refilling and battery-charging services) are out of scope. A future revision may extend the model to cover services if the field matures in that direction.
9. Closing
The cannabis consumption hardware safety field is at the moment when its certification mechanism gets specified. The temptation will be to import the binary mark as it exists in adjacent industries, because it is administratively cheap and visually familiar. We argue that the temptation should be resisted. The field is heterogeneous enough, and the science young enough, that the appropriate certification idiom is provenance — what was tested, by whom, and under what protocol — surfaced via a tier-graded mark and backed by a public registry.
The CDSI three-tier model is one specification of that approach. We do not believe it is the only viable specification; we believe it is a defensible starting point that respects the field’s heterogeneity, admits resource-constrained participants without diluting the meaning of the highest tier, and preserves the institutional credibility on which any voluntary certification ultimately depends.
The tier structure is intended to be revisited. The acceptance criteria, the re-test cadences, the fee structure, and the visual marks are all governance artifacts that will evolve as the field’s analytical methodology matures and the participant base scales. We are publishing the first specification so that other groups — including standards bodies (ASTM, UL, NSF), other safety institutes, and regulators — can engage with it on its merits and improve on it.
The mark on the device is the visible artifact. The registry behind the mark is the substance.
References
European Union. Commission Regulation (EEC) No 2568/91 of 11 July 1991 on the characteristics of olive oil and olive-residue oil and on the relevant methods of analysis. Official Journal of the European Communities, L 248, 1991.
International Organization for Standardization. ISO/IEC 17025: General requirements for the competence of testing and calibration laboratories. 2017.
Macosko, M. The Hardware Vacuum: Why Cannabis Consumption Devices Have No Safety Standard. CDSI Working Paper P-001. 2026.
Macosko, M. Loaded-State Off-Gas Analysis of Cannabis Concentrate Vaporizers. CDSI Working Paper P-002. 2026.
Macosko, M. The Pyrolytic Reservoir: Why Single-Draw Testing Underestimates Real-World Cannabis Vaporizer Hardware Risk. CDSI Working Paper P-003. 2026.
NSF International. Certification Programs Overview. https://www.nsf.org/standards-development/certification.
Underwriters Laboratories. Standard for Electronic Cigarettes and Vaping Devices. UL 8139, 2018.
U.S. Environmental Protection Agency. ENERGY STAR Program: Most Efficient Designation Methodology. 2011 (and subsequent annual revisions).
Working paper. Comments to matt@ineedhemp.com (papers@cdsi.click pending provisioning). This draft is v0.4 (May 2026); the next planned revision is concurrent with the public release of CDSI-002 (operationalizing protocol).