While AI is currently a popular topic, its value in the transport industry is defined by its ability to save time, improve safety, and increase margins rather than just being a "flashy" tool. Allotrac’s AI tools are being developed to solve specific bottlenecks, such as reducing administrative costs and recovering lost time, allowing personnel to manage more assets and jobs without burning out.
While Allotrac is not primarily an AI company, our roots as a Wollongong-based Transport Management System span more than ten years of industry success.
Our suite of AI-driven features is constructed upon this foundation of field-tested technology. The following sections explore how these advancements provide tangible commercial benefits to the transport sector.
Phase 1: Conversational AI with "Allie"
The first phase of this AI integration is a conversational chat model called Allie. Allie is embedded directly within Allotrac.io to assist teams with various tasks:
- Instant Data Retrieval: Users can ask natural language questions, such as comparing current average wait times to the weekly average, to get instant answers without digging through complex reports.
- Back-Office Support: Allie can handle repetitive administrative requests, such as gathering Proof of Delivery (POD) documents from the last week into a single PDF and emailing them to a specific address.
- Operations Summaries: Users can prompt the AI for an end-of-day summary of operations and completed tasks.
- Managing the Unexpected: In the event of a truck breakdown, Allie can automatically reallocate loads to another truck within the TMS and notify the new driver, reducing the manual coordination typically required of dispatchers.
Phase 2: Automated Workers
The second phase moves from a reactive chatbot to autonomous team members that operate in the background 24/7 without needing to be prompted. These automated workers are defined by:
- Job Descriptions:Defined Roles: Each agent is assigned a unique function, such as a "weighbridge reconciliation agent." This worker manages supplier statements by validating them against TMS records. If verified, the agent forwards the data to accounts for payment; should a discrepancy arise, it is automatically escalated to the relevant manager for resolution.
- Event-Based Triggers: Workers are activated by specific actions, such as a job status change, an inbound email, or a phone call.
- "Hands" for Execution: Agents are given the "tools" needed to complete their jobs, including access to update data within the TMS or utilize outbound communication services like SMS and email.
Read more on Automated Workers.
Safety and Security
To ensure safety and prevent "hallucinations," these agents operate within a strict security and permissioning system. They are only allowed to use the specific data they have been granted access to and must follow precise instructions on how to respond.
Real-World Application: Autonomous Communication
Modern autonomous agents can now handle complex tasks like making phone calls to customers. For example, an agent can be programmed to call a customer once a job is allocated to confirm delivery details, addresses, and quantities, and even update the order if the customer requests changes.
Ultimately, the goal of these AI tools is not to replace humans, but to automate mundane, repetitive tasks, allowing team members to focus on revenue-generating activities.
.png)