An AI agent can create. It can predict. It can analyze data. But for it to really drive change and enhance operations, it needs to take action. That’s the next step of machine learning and AI, says Kence Anderson, CEO of Composabl, a company that helps manufacturers create multi-agent AI systems.
Our discussion with Kence is probably one of the most important you’ll hear on the podcast lately. AI has been a big topic for a while now, but we dig further into how it can be applied at an industrial level to help manufacturers and engineers make decisions and solve problems.
We start with Kence’s overview of AI agents, specifically in manufacturing, and talk about what drew him to this industry. Kence talks about how companies can find a balance between capturing data and also using expertise from your team (and why it’s crucial to do this before they start retiring!)
Subscribe
Apple | Google Podcasts | Spotify | Stitcher | YouTube
3 Quick Things from this Episode
- Creating simulations for AI agents to take a task on a practice run before you go live with it is a great way to iron out any issues and test out the best deployment methods. These simulations can teach you more and also train the AI in the process.
- Quantifiable data has been a big focus for manufacturers and engineers over the past few years because of what it can teach us and help us predict. But it’s also important to look beyond the numbers and interview your team. They’re experts and could tell you a solution 10x faster than an AI agent would take to learn that from the data it analyzes.
- AI agents work best when trained with highly specific information. This doesn’t restrict them because an AI agent doesn’t need to do absolutely everything. It just needs to be really good at solving those specific problems your team faces.
Enjoying the show? Please leave us a review here. Even one sentence helps. It’s feedback from Manufacturing All-Stars like you that keeps us going!
Tweetable Quotes
- “ The thing that moves the needle is decisions. And so, if you’re predicting something, analyzing something, it’s only to serve the purpose of making great decisions.”
- “ You can have an AI just look at your data and discover a bunch of stuff that you know Susan and Joe over there could have told you in 15 minutes. So there’s this need to capture and codify high-value subject matter expertise, frankly, before it goes extinct.”
- “There’s a hundred million engineers out there in an industry that if you empower them with the right tools, they’ll innovate and engineer systems the same way as they have everything else.”
Please Scroll Down for Recommended Resources from this Episode
Do you want to connect with other leaders that are moving the needle in manufacturing everyday? Then make sure to join us in the Manufacturing Happy Hour Industry Community on LinkedIn.
Skillwork is the premier staffing agency focusing exclusively on skilled labor. Their team of highly skilled workers augments existing teams for elevated impact and decreased downtime. Partnering with hundreds of facilities in over 30 states, Skillwork deploys their traveling industrial technicians to work for you on a contract basis.
Layerly creates professional websites for small businesses and manufacturers in just 5 days, all managed under a simple subscription. With built-in SEO and personalized support, they’re here to help you succeed ASAP—no hassle, just results.
What We Discuss with Kence
- What is an AI agent and how could manufacturers implement them to take action?
- Ways AI agents can take how engineers and manufacturers already use machine learning to the next level
- Kence explains how he first entered the manufacturing world with a mechanical engineering degree and roles in startups
- Why Kence believes that high-value technology needs methodology combined with a platform to enable widespread adoption
- 3 reasons why AI in manufacturing has always fascinated Kence
- Why AI agents need to be taught specific information and why this enhances rather than restricts how they work
- The importance of collecting data but also interviewing experts in your team to gain their unique knowledge
- Kence shares an example of how an extruder making Cheetos is way harder than just programming a machine
- How Kence uses simulations to enable AI agents to practice and learn tasks before they’re implemented
- Where does the AI agent live? Kence gives a detailed explanation of how AI agents are stored and communicate with wider internal systems
- Different ways AI agents can not only be taught but also teach others
- The steps manufacturers need to take to start using AI agents and why the first one should be to identify the high-value skills in danger of extinction
Recommended Resources
- Composabl, a platform that combines the power of Composite AI with expertise to design and deploy multi-agent AI systems for industrial environments
Recent Comments