Frequently Asked Questions (FAQs)

Given that we have to configure the platform, integrate with your internal systems, load data and create custom ai agents, there will be work and coordination between Spector and your business to set up a working system. Additionally, AI Agents are a new concept so it will require us to train your team on how to use the platform and maximize their benefit from it.

While the duration of the onboarding process will depend on the scope of the project, typically it takes about two weeks to get familiar with customer data (background documents related to assets, maintenance history, standard practices, etc.) and an additional one to two weeks to ingest customer information, extract knowledge, train and setup the required models.

An agent is an entity capable of agency, i.e. independent decision making and action. Autonomy or even partial or delegated autonomy requires contextualized situation awareness, supported by sensing (making observations/measurements) and perception (inferring the state of the world from observations), causal modeling of the world (actions from agents, including one’s own, and other possibly non-agentic latent causes bring about predictable changes in the world), and the understanding of one’s own agency: the space of actions possible with their predicted impact.

An AI Agent is an intelligent ‘computer’ system capable of sensing, perception, causal world modeling, and autonomous action.

A Copilot is a communicative agent capable of assisting a human on pre-specified tasks (that the agent is trained on). For effective assistance, the agents is able to (a) undertake sophisticated two-way communication and dialog, (b) interpret human intent to understand the task and the associated sub-tasks and actions that it needs to execute, (c) create plans for executing the task including being able to dynamically break it down into multiple steps, (d) invoke backend tools and skills (both AI or non-AI) to execute the individual steps, (e) collate and provision the results back to the human supervisor, (f) operate in a ‘delegated autonomy’ mode to the human, explaining its plans and results, proactively providing available options to the human supervisor and seeking assistance from him whenever needed.

Spector’s agents and Copilots are created for asset health management. Our agents are co-pilots to reliability engineers, allowing them to focus on their primary tasks, while the Copilot is able to take on the most data-intensive tasks, thereby significantly increasing the efficiency of the reliability engineer.

In particular, our RE (Reliability Engineer) Copilots assist with the following RE responsibilities:

• Preventative Maintenance
• Predictive Analysis
• Root Cause Analysis (RCA)
• Performance Evaluation Assessments
• Overall Equipment Effectiveness (OEE) monitoring
• Conducting Failure Mode and Effects Analysis (FMEA)
• Developing Maintenance Procedures and Schedules
• Creating Life Cycle Asset Management (LCAM) plans
• Reviewing Newly Installed Plants and Equipment
• Assisting in Design and Development Planning for Custom Assets