@CQB45ACP - much of my career has been in manufacturing process innovations, which inevitably includes automation technology projects such as what you describe.
Much of those systems would already be quite automated - mixers, batch weighers and bulk loaders, dryers, etc. However, knowing a bit about the chemical composition and therefore speculating the margins, I can relate my own experiences in manufacturing technology value engineering - “robotics” as a blanket solution is a very 1970s & 1980s-esque vision. Post-WWII industrial revolution and the era of the Jetson’s really threw folks into a mindset of “robots will solve everything,” chased by the renovation of Detroit and auto manufacturing. And naturally, sci-fi predictions running ahead of the industry with movies like the Terminator series in which automation innovation ran so quickly we couldn’t even keep up as a species…
But reality in industry is much less sexy… As an example, MOST manufacturing automation projects which are proffered never come to realization, simply because the market will not tolerate the extreme product cost increases which must accompany such process upgrades. As a scalar, most projects don’t move forward if payback isn’t realized within 2 years - whereas some of the manual labor risk displacement projects (“robotics projects”) I have had proposed to me have ended up with 30-80 year paybacks… meaning they literally would never payback, as the life of the machines, life of the products they’d make, and even the MARKET for the products weren’t long enough to ever pay back the investment and operation cost. I led a project a few years ago which would replace 140 workers (of which we could only reliably staff about 65-80 at any time, meaning the production targets were being missed and EXTREME overtime costs were being paid). Even overpaying relative to the regional market by 30%, we couldn’t fill our production floor… so our robotics project totaled over $185 million, and would have taken over 20yrs to payback, with the robots needing maintenance effectively re-buying the project every 10yrs… some of the project survived, largely in a pack & stack portion of the plant, but the rest of the project was scrubbed… cheaper to pay double overtime on top of 30% over market wages on labor than to buy the robots to replace them…