Dubai RTA Has 40 AI Products in Its Strategy. Only 11 Are Complete So Far

Dubai's Roads and Transport Authority has moved its artificial-intelligence programme beyond a collection of individual experiments and into a multi-year operating strategy.

The updated plan spans 41 areas across RTA agencies, sectors and departments.

Its first two rollout waves cover 11 business areas and include 40 AI products.

Eleven of those products have already been completed.

Eleven products are operating evidence. Forty is the programme

RTA says it has identified more than 180 AI use cases and quick-win initiatives across the first 11 business areas.

Those use cases have been organised into 40 products, with 11 launched so far.

That distinction matters because a strategy containing 40 products should not be reported as 40 completed AI deployments.

AED300 million is a target, not a realised annual return

Over the strategy's two-and-a-half-year implementation period, RTA aims to generate more than AED300 million in annual added value.

It also aims to move 50 percent of operations and services into an Agentic AI ecosystem.

Those figures describe the intended destination of the programme.

RTA has not said that the full AED300 million annual value or 50 percent transformation level has already been achieved.

Some operating results are already available

RTA says active usage of Microsoft Copilot exceeds 90 percent among employees who have access to it.

The authority reports a 10 percent saving in working time among those users.

That provides a current operating indicator that is different from the larger future targets attached to the strategy.

Mahboub is moving from answering questions to completing transactions

RTA is also upgrading its Mahboub virtual assistant using Agentic AI.

The new version can complete five vehicle-licensing services and provide approximately 100 information services.

Integration with UAE PASS and DubaiPay allows supported vehicle transactions to be completed through the same conversational journey.

RTA says the system aims to increase the share of services delivered through Mahboub to 10 percent by the end of 2026 and 50 percent by the end of 2027.

The eNOC system has a large existing workload to work against

One of the strategy's Agentic AI products is an electronic No Objection Certificate system for works and projects near RTA assets.

RTA says the service currently issues approximately 130,000 certificates annually.

The AI system is intended to guide applicants, perform preliminary document checks and provide engineering insights to approving entities.

Its performance goals include accelerating review procedures threefold and improving customer satisfaction and revenue.

An AI auditor is targeting full document coverage

RTA's Licensing Agency has also developed an Agentic AI Auditor that uses machine learning and optical character recognition to review licensing documents.

The authority says the system is intended to move from sample-based auditing to 100 percent of documents within its defined scope.

It is also targeting audit processes up to 21 times faster and more than 25,000 hours of manual work saved annually.

Those remain programme targets until measured performance is published.

Twenty AI agents are being applied to bus planning

RTA's Network Planner uses approximately 20 AI agents to support analysis of Dubai's bus network.

The system evaluates performance, identifies service gaps, simulates alternatives and recommends network changes.

RTA is targeting faster analysis, shorter implementation time for service changes and improvements in punctuality and cost recovery.

What Arabian Current is watching

Watch how many of the remaining 29 products move into operation, which performance targets are independently measurable and whether the 50 percent Agentic AI target is reached within the programme period.

The strongest future evidence will come from comparing promised savings with actual operating results.