Fleet360 AI FAQ: Ron360, Ron360 Reports, FleetHQ, and AI credits

What AI features are in Fleet360?

"RonF360", "Ask Ron", and "Ron 360" all refer to the same assistant: Ron360.

Pricing and allocation

How is AI priced in Fleet360?

AI usage is metered in AI credits, where 1 credit = 1 US dollar of AI usage. Every account receives a monthly plan-included allowance of credits. Usage is drawn from that allowance first; if the allowance is used up, the account can purchase additional credits (see What happens when we run out?).

Is the allowance shared across the account or allocated per user?

Shared. The allowance is a single pool for the entire account. Every user's AI usage, across every AI feature, draws from the same pool. There is no per-user quota, although administrators can see a per-user breakdown of who is consuming credits.

Does the allowance roll over?

No. Plan-included credits reset to the monthly allowance on the 1st of each month; unused plan credits are discarded. Purchased ad-hoc (paid-extra) credits never expire and are spent only after the plan-included balance reaches zero.

Is the allowance an initial setup / development cost, or ongoing usage?

It is an ongoing monthly usage allowance, not a one-time setup fee. There is no separate charge to "set up" Ron360, Ron360 Reports, or FleetHQ; credits are consumed only when the AI actually does work (answering a question, building a report, researching an insight).

How usage is calculated

What determines how many credits a request uses?

Credits are calculated from the AI model work required to answer the request, measured in tokens (input, cached input, and output) at the rate for the model used, plus RTA's pricing multiplier. It is not based on the amount of Fleet360 data returned, the number of rows in a table, or the number of users.

In practice:

• Simple questions cost less; complex questions that require multiple queries, larger context, or charts cost more.

• A typical Ron360 question costs roughly $0.10–$0.20. Actual cost varies with complexity and the model in use. Other use cases (like Ron360 Reports) use smarter models, and therefore cost more.

• Follow-up questions in the same conversation carry the conversation context, so long conversations gradually cost more per turn. Start a new conversation when changing topics.

• Knowledge-base-only answers (how-to / help questions) are not charged.

Are credits charged per query, per run, or per report?

Is there an additional cost each time a saved report is run?

No. Once a Ron360 report has been created, running it, exporting it, and scheduled deliveries do not consume AI credits. Credits are consumed only when the report is built or edited with AI. Existing reports continue to run even if the account's AI credits are exhausted.

How is a failed or cancelled request handled?

Fleet360 reserves a small amount of credit when a request starts and reconciles to the actual usage when it finishes. If the request fails or is cancelled, the unused reservation is refunded.

Access and permissions

Is access determined by user roles / permissions?

Yes. Each AI feature is gated by permissions, so a customer can roll AI out to a small group first and expand later:

The AI features also require AI to be enabled for the account, and the relevant system switches (for example the AI chat assistant switch and the FleetHQ switch). If a user does not see the Ask Ron360 button or the FleetHQ / Ron 360 Reports menu items, check the account-level enablement first, then the user's permissions.

Can we control spend by limiting who has access?

Yes. Because the allowance is shared, the most direct cost control is limiting which users hold the AI Use, AI Reports Add/Edit, and Act on FleetHQ Insights permissions. Users with only AI Reports View can still run and export existing Ron360 reports without spending credits.

Does Ron360 see data a user is not allowed to see?

No. Ron360 answers only from the data and facilities the signed-in user is permitted to see, and applies the same field-level redaction as the rest of Fleet360. A shared Ron360 report runs against the recipient's own data access, so results can differ from the owner's. Scheduled reports sent to external email addresses are generated with the schedule owner's access; treat delivery links as sensitive.

Monitoring and running out

How can we see how much we have used?

Admin → AI Consumption. The Usage tab shows the month's credits spent, remaining, and request count, with a per-user breakdown (click a user for their conversation-level detail). The Usage Management tab shows the plan-included balance, paid-extra balance, total available, grant/transaction history, and auto top-up settings. A month selector allows review of prior months.

What happens when we run out of credits?

• Ron360 chat: new questions are blocked and the chat panel shows a message to contact an administrator.

• Ron360 Reports: new builds and edits fail with "AI credit limit reached, contact your administrator." Existing reports continue to run, export, and deliver on schedule.

• FleetHQ: insight detection continues; research that needs credits pauses until credits are available.

Credits are restored automatically on the 1st of the next month, or immediately when additional credits are added.

How do we get more credits?

1. Purchase additional credits by emailing addons@rtafleet.com. Purchased credits are added to the paid-extra balance and never expire.

2. Increase the monthly allowance by adjusting the monthly subscription by emailing addons@rtafleet.com.

Can FleetHQ spend credits without anyone asking it to?

Only if the FleetHQ automatic research system switch is on (it is off by default). With it off, FleetHQ still detects and reports insights, and an authorized user can request research on a specific insight manually. Turn automatic research on only when the account is comfortable with FleetHQ spending credits on a schedule.

Data, accuracy, and trust

Is our data used to train AI models?

Fleet360 sends the minimum context needed to answer a request to the AI model provider under commercial API terms. Customer data is not used to train the provider's public models

How accurate is Ron360?

Ron360 answers from live Fleet360 data and shows its results as prose, charts, or tables so they can be checked. Like any AI assistant it can misinterpret an ambiguous question. Best practice: include the business area, metric, time period, and grouping ("maintenance spend by facility, year to date"), ask follow-ups to refine, and use the thumbs up/down controls so RTA can improve answers. Ron360's help/how-to answers are not authoritative documentation;

Quick answers for the common questions