Direct Answer
The platform must give group headquarters full visibility, yet let an outsourced electrician see only one workshop — how should permissions be divided? Authorize in three dimensions: "role × data scope × operation level" — the role defines the function (view/confirm/handle/configure), the scope defines the visible topology (group/campus/workshop/single panel), and the level defines whether parameters can be changed; outsourced accounts are read-only plus alarm confirmation by default. The multi-role permission design of the FEXCloud IoT cloud platform: four roles (owner/operations/supervision/outsourced) each matched with "what to see, what to do, whether to change" — combined in three dimensions, permissions never cross the line.
A scene: four kinds of people on one platform: group headquarters needs full visibility, campus operations needs to handle alarms, regulators need to audit data, and an outsourced electrician is responsible for only one workshop — if every account has one permission, headquarters cannot see the details and the outsourced worker can change global parameters — the essence of permission design is not "whether to give an account" but "what each role sees, what it can do, and whether it can change" — authorized in three dimensions, the four roles each in their place.
1. The Essence of Permission Design: Three-Dimensional Authorization
Dimension one: role — the role defines the "function boundary": is this account an owner, operations, supervision or outsourced — the role decides which function modules can be called; Dimension two: data scope — the scope defines the "visibility boundary": group/campus/workshop/single panel — the scope decides which equipment and data can be seen; Dimension three: operation level — the level defines the "operation boundary": view/confirm/handle/configure — the level decides what actions can be taken; The three-dimensional combination — any account = role × scope × level: an outsourced electrician = outsourced role × workshop scope × read-only level — after the three-dimensional combination, the permission is precise and never crosses the line.
In one sentence: the three dimensions of permission = role defines the function + scope defines the visibility + level defines the operation — combined in three dimensions for precise permission — the era of "one administrator account" is over.
A three-dimensional example: four accounts at a campus: headquarters (owner × whole group × view), campus (operations × this campus × handle), supervision (supervisor × full scope × read-only audit), outsourced (outsourced × workshop × read-only + confirm) — four permissions on one platform, none crossing. Four accounts, four "three-dimensional coordinates" — role, scope and level each set one dimension — different coordinates, different permissions.
2. RBAC Role Design: The Four Roles
Role one: owner (group/campus management) — sees operating reports, energy overviews and compliance status — focuses on "the overall situation and trend" — view level by default; Role two: operations (duty/engineer) — sees equipment data, confirms alarms and initiates handling — focuses on "operating state and anomaly handling" — confirm + handle level by default; Role three: supervision (safety/compliance department) — sees audit records, compliance reports and permission logs — focuses on "traceability and compliance" — read-only audit level by default; Role four: outsourced (third-party service provider) — sees only the authorized equipment and confirms alarms assigned to it — focuses on "the part they are responsible for" — read-only + confirm level by default.
In one sentence: the four roles = the owner sees the whole + operations manages handling + supervision checks compliance + outsourced does its share — the role defines the function, and with the function boundary drawn first, permission has a landing point.
A role example: account requests on a platform: group finance needs energy reports (owner role), an electrical engineer needs to handle alarms (operations role), the safety-environment department needs to export audit logs (supervision role), and an outsourced electrical team needs to receive its workshop's alarms (outsourced role) — four requests, four roles, each taking what it needs. A role is not a title but a function boundary — define the role first when requesting an account, and the function boundary is drawn first.
3. Data Scope Bound to Topology
The levels of scope — the data scope binds to the electrical topology: group (all campuses) → campus (this campus) → workshop (this workshop) → single panel (this panel) — the scope narrows level by level; How the scope binds — the account scope = topology-node authorization: authorizing "a campus" means seeing all equipment under that campus; authorizing "a workshop" means seeing only that workshop's equipment; The inheritance of scope — an upper scope covers the lower by default: the group scope sees all campuses, the workshop scope sees only this workshop — the inheritance rule keeps authorization clear without duplication; The role of scope — the scope decides "data visibility": an outsourced electrician cannot see other workshops' data — visible means permitted, scope means boundary.
In one sentence: the data scope binds to the topology = group/campus/workshop/single panel narrowing level by level — an account authorized to a node sees that node's data — scope is the boundary, and visibility is the permission.
A scope example: an outsourced electrician account at a campus: authorized scope = workshop 3 panels — after login, only workshop 3 equipment and alarms are visible — other workshops' data is simply invisible in the interface. Scope is authorized to the node — authorized where, seen where — the outsourced electrician sees nothing beyond the workshop, and permissions naturally never cross.
4. Operation Levels: View/Confirm/Handle/Configure
Level one: view — read-only: see data, reports and alarms — no operation — for supervision and read-only scenarios; Level two: confirm — confirm alarms (mark "noted/in handling") — cannot change parameters — for duty confirmation; Level three: handle — confirm + initiate handling actions (issue commands, reset, linkage test) — cannot change system configuration — for operations engineers; Level four: configure — handle + modify system configuration (thresholds, model parameters, role permissions) — the highest permission — for system administrators; The boundaries of the levels — the levels progress and cannot skip: the configure level can do everything, the view level can only look — outsourced defaults to read-only + confirm, naturally never touching configuration.
In one sentence: the four operation levels = view (only look) + confirm (mark) + handle (action) + configure (change parameters) — the levels progress without skipping — outsourced defaults to read-only plus confirm, naturally never touching configuration.
A level example: a comparison of three account operations on a platform: supervision (view level: can only see audits and reports), operations (handle level: can confirm alarms and issue reset commands, but cannot change thresholds), administrator (configure level: can modify model parameters) — the levels are clear and operations do not cross. The level defines the operation boundary — those who can see cannot necessarily act, and those who can act may not change — four clear levels, permissions never cross.
5. The Least-Privilege Principle for Outsourced Accounts
The meaning of least privilege — an outsourced account gets only "the minimum permission needed to complete the work": the smallest authorized scope and the lowest operation level — just enough, with nothing extra; The default configuration for outsourced accounts — read-only + alarm confirmation: can see the alarms assigned to it and confirm the handling state — cannot change parameters or see other scopes; The value of least privilege — outsourced workers are third parties with high personnel turnover and high account risk — least privilege minimizes "the impact an outsourced account can cause"; The dynamic management of outsourced permissions — revoked when the project ends, changed when personnel changes — outsourced account lifecycles are short, and permission management must keep up.
In one sentence: outsourced least privilege = smallest scope + lowest level + dynamic revocation — outsourced accounts are read-only plus confirmation, revoked when the project ends — third-party risk is naturally reduced by least privilege.
An outsourced-permission example: an outsourced electrician account at a campus: read-only + alarm confirmation, scope limited to workshop 3, valid for 6 months — the account was revoked when the project ended — the departure of outsourced personnel did not affect platform security. An outsourced account is "just enough" — the smallest scope, the lowest level and revocation on expiry — third-party risk is kept out by the permission design.
6. Permission Auditing and Regular Review
What auditing covers — records the account operation log: who saw what, who changed parameters, who exported data — operations leave traces and responsibility is traceable; The focus of auditing — high-risk actions are the priority: configuration changes, data exports and permission changes — high-risk actions always have logs; Regular review — permission reviewed regularly: is the account still needed, is the scope still accurate, is the level still appropriate — review quarterly or semi-annually; The handling of the review — the review finds "departure without revocation, scope too large, level too high" → adjust immediately — the review is the "physical checkup" of the permission system.
In one sentence: permission auditing and review = operations leave traces + high-risk actions audited first + regular review + fix problems immediately — auditing keeps permissions traceable, and review keeps them from expiring.
An audit example: a quarterly review on a platform found: 3 accounts of departed staff not revoked, 1 outsourced account wrongly configured to the whole campus, 1 operations account holding the configure level — all adjusted immediately. Permission is not "set and done" — operations leave traces, review runs regularly, and problems are fixed immediately — a healthy permission system is what keeps security guaranteed.
7. The Three-Dimensional Combination Example: A Comparison of the Four Accounts
The group-headquarters account — owner role × group scope × view level: sees the group's energy and compliance status, does not join single-point handling — the executive perspective; The campus-operations account — operations role × campus scope × handle level: confirms alarms and issues handling commands — the front-line perspective; The supervision account — supervision role × full scope × read-only audit level: exports audit records and checks compliance — the compliance perspective; The outsourced account — outsourced role × workshop scope × read-only + confirm level: receives its own alarms and confirms the handling — the in-scope perspective.
In one sentence: the four accounts are four sets of coordinates — the owner sees the whole board, operations manages the front line, supervision checks compliance, and outsourced does its share — configured once in three dimensions, permissions never cross.
A combination example: four kinds of accounts online on one platform at the same time: headquarters viewing the group report, campus operations handling this campus's alarms, supervision exporting audit logs, and outsourced confirming workshop 3's alarms — one platform, four permissions, no interference. The three-dimensional combination lets "one platform, four kinds of people" each take what they need — role, scope and level each set one dimension, and permissions naturally do not conflict.
8. A Complete Configuration Example
The full flow: a permission configuration for a group platform: ① role definition — owner (group + campus), operations (duty + engineer), supervision (safety-environment), outsourced (third party) → ② scope binding — headquarters = group, campus operations = this campus, supervision = full, outsourced = single workshop → ③ level setting — headquarters = view, campus operations = handle, supervision = read-only audit, outsourced = read-only + confirm → ④ audit configuration — high-risk actions leave traces end to end → ⑤ quarterly review — revoke departures, calibrate scopes, adjust levels. From role to review: define the role first, then bind the scope, then set the level, configure auditing and review regularly — with the five steps done, the multi-role permission system lands.
In one sentence: the permission configuration flow = define roles + bind scopes + set levels + configure auditing + review regularly — with the five steps done, the owner, operations, supervision and outsourced are each in their place.
9. Common Mistakes in Permission Configuration
Mistake one: one account for everything — an "administrator account" shared by everyone: anyone can change configuration and export data — when something goes wrong, no one can be identified — the right way: one account per person, authorized by role; Mistake two: roles only, no scope — all operations accounts have the same permission: outsourced workers can see the whole campus's data — without scope bound to topology, the role is meaningless — the right way: authorize every account to a specific topology node; Mistake three: outsourced accounts equal to internal — outsourced workers can handle any alarm and modify parameters — a third-party risk exposure — the right way: outsourced defaults to read-only + confirm, with the smallest scope; Mistake four: set and forget — accounts only grow: no revocation on departure, no adjustment on transfer — permissions pile up ever wider — the right way: review quarterly and fix problems immediately; Mistake five: audit logs as decoration — logs stored but never checked: anomalies are never found — auditing loses its meaning — the right way: high-risk actions checked first, anomalies traced immediately.
In one sentence: the five permission mistakes = shared accounts + scope not bound + outsourced equal rights + only growth without reduction + auditing not checked — five mistakes, five pits; the permission system is healthy only when they are avoided.
A mistake example: a campus once shared "one administrator account" across the whole department — the account was not revoked after an outsourced worker left — one misconfiguration triggered false alarms across the campus — investigation found that the shared account made it impossible to identify the responsible person. Every permission incident comes from a "cutting corners" configuration — no shared accounts, outsourced minimized, review routine — and incidents naturally decrease.
10. Back to FEXLINK: How the Multi-Role Permission of FEXCloud Lands
The FEXCloud IoT cloud platform has built-in "role × data scope × operation level" three-dimensional authorization: it supports custom roles, topology-node authorization, four operation levels, operation auditing and regular review, with visual permission configuration and traces for every change. The configuration checklist: ① define the roles first (owner/operations/supervision/outsourced) → ② bind the data scope to the topology (group/campus/workshop/single panel) → ③ set the operation level per role (outsourced defaults to read-only + confirm) → ④ audit high-risk actions end to end → ⑤ review permissions quarterly (revoke departures, calibrate scopes) — for multi-role permissions, the three-dimensional design lands at once.
One summary: multi-role permission design = role defines the function + scope defines the visibility + level defines the operation — the owner sees the whole, operations manages handling, supervision checks compliance and outsourced does its share; outsourced accounts default to read-only plus alarm confirmation, and the least-privilege principle runs throughout. Permission is not a question of "whether to give an account" but of "what each role sees, what it can do and whether it can change" — designed in three dimensions, the four types of accounts each in their place.
One reminder: when a new account is requested, do not open it right away — answer three questions first: is the role an owner, operations, supervision or outsourced? To which topology node does the scope reach? Is the level view, confirm, handle or configure? — with the three answered, the permission is automatically precise. Every permission crossing started with a question that was not asked — ask the three dimensions first, and crossings do not happen.
Scope and Limitations
- The conclusions in this article are intended for LV distribution and electrical-safety sites. Criteria are compiled from public codes and FEXLINK product documentation, and product parameters follow the latest selection handbook and product knowledge base.
- Quantitative criteria (protection zone, discharge and residual voltage, earth resistance, thresholds and grading) hold under the preconditions stated in this article. Where the site voltage level, earthing arrangement, environmental category or load profile differs, the figures must be recalculated. Specific configurations involving FEXCloud IoT cloud platform should be verified against actual circuit parameters.
- Specific product models and parameters are subject to the latest selection handbook and product knowledge base; this article does not constitute selection or procurement advice.
- Standard numbers cited here refer to the currently effective editions; when a standard is revised, the related quantitative conclusions must be re-checked.
- Deployment and implementation for specific projects must be carried out by qualified engineers based on site survey and design drawings.
Frequently Asked Questions
Designing Multi-Role Access
The platform must give group headquarters full visibility, yet let an outsourced electrician see only one workshop — how should permissions be divided