One of the bigger questions surrounding the use of AI (artificial intelligence) and associated technologies like ChatGPT is: Can this new, transformative technology be used to free us up to focus on more high-level strategic tasks?
At Huffman Engineering, we’ve been dabbling with the use of AI to see if we can find workable use cases to deliver greater efficiencies in our work. In the process we’ve come across a few potential use cases we think might be viable trials for use in industrial control systems integration.
One of these use cases involves the use of ChatGPT to generate an outline of the steps required to design the complex systems we implement. The idea is to use these steps to guide a new engineer through the process.
Here’s the ChatGPT prompt we used: “Designing a three-pump lead-lag lift station for an industrial automation engineering application involves careful consideration of various factors such as system requirements, redundancy, safety and efficiency. Here's a general outline of the design process:”
And here is the result from ChatGPT:
1. Determine System Requirements:
- Calculate the flow rate and head requirements for your lift station based on the specific application.
- Identify the desired level of redundancy and reliability.
2. Pump Selection:
- Choose three pumps that are suitable for the application, taking into account factors like flow rate, head, efficiency and material compatibility.
- Ensure that the pumps can handle the potential variations in flow and head.
3. Control System:
- Implement a control system that manages the lead-lag operation of the pumps. A programmable logic controller (PLC) or a dedicated pump controller can be used.
- Configure the control system to monitor pump performance, start and stop pumps as needed, and provide alarms for faults or irregularities.
4. Redundancy and Reliability:
- Design the system for redundancy to ensure continuous operation even if one pump fails. This typically involves having one lead pump and two lag pumps.
- Implement an automatic pump rotation schedule to distribute wear and tear evenly among the pumps.
- Include backup power supply options (e.g., generators or uninterruptible power supplies) to maintain operation during power outages.
5. Safety Measures:
- Incorporate safety features such as high-level alarms, low-level alarms, and overflow prevention mechanisms.
- Ensure that the system complies with relevant safety standards and regulations.
6. Monitoring and Remote Access:
- Integrate a remote monitoring and control system that allows operators to monitor the lift station's performance and receive alerts remotely.
- Implement data logging and analytics to track pump performance over time and predict maintenance needs.
7. Installation and Maintenance:
- Ensure that the lift station is installed correctly, and all components are properly maintained.
- Establish a regular maintenance schedule to inspect pumps, motors, controls and other components.
8. Instrumentation and Sensors:
- Install sensors to measure parameters such as flow rate, pressure and pump status.
- Use level sensors in the sump to control pump operation based on liquid levels.
9. Emergency Response Plan:
- Develop an emergency response plan that outlines procedures for handling pump failures, power outages or other critical situations.
- Train personnel in these procedures.
10. Documentation: Maintain detailed documentation of the lift station design, including schematics, control logic diagrams and maintenance records.
11. Testing and Commissioning: Before putting the lift station into full operation, perform thorough testing and commissioning to ensure that all components and control systems work as intended.
12. Regulatory Compliance: Ensure that the lift station complies with all relevant industrial and environmental regulations.
Remember that the specific details of the design will depend on the unique requirements of your industrial automation engineering application and local regulations. Consulting with experts in industrial automation and wastewater management is essential to create a safe and efficient lift station.
A cautionary tale
While these results demonstrate that AI can provide viable options for improving workload efficiencies, there are also concerns around using AI in everyday engineering which warrant serious policy and procedural considerations, such as:
- A new person to the industry dependent solely on AI or autogenerated program designs will never learn the ins and outs of the industry. Relying largely on AI-generated directives will never provide a full understanding of what it takes to design and build an entire system.
- An unseasoned professional could also learn to blindly trust what is given to them as truth. Trusting a design that may not be well thought out could prove to be a costly mistake.
- Perhaps the biggest danger is in the potential for a loss of creativity that could result in engineers developing a dependence on technology to design systems as a shortcut. Forgetting to lean on their true intelligence as an engineer to help manufacturers navigate challenging obstacles and achieve great efficiencies, higher productivity or safer systems for their companies is a risk no system integrator should take.
Keith Mandachit, PE, is engineering manager at Huffman Engineering Inc., certified members of theControl System Integrators Association(CSIA). For more information about Huffman Engineering, visit its profile on the Industrial Automation Exchange.
