________________________________________________________________________________________________________________________
2 . What are the current limitations in warehouse automation ? Historically , automation has been good for handling single , standardized product categories but terrible at dealing with anything more complex . The problem for third-party logistics ( 3PL ) providers is that if they rely on this type of crude automation , they end up turning away every client whose products don ’ t fit the bill .
Worse still , many of the big 3PLs have centered their day-to-day warehouse operations around robots and machines with limited functionality , effectively making their human team members subservient to them . This leads to meaningless jobs that no one wants to do , hence why the warehousing industry churn rate is so high .
3 . More specifically , can you discuss the integration of AI and automation within warehouse logistics ? Until recently , every different automation use case has required a different robot or , at the very least , different tooling . With AI , the situation changes radically . It makes it much easier to build humanoid robots - or similar - that serve multiple use cases , and crucially , the tech will ensure the robots can adapt in real time to anything that doesn ’ t fit the brief .
Based on the current speed of AI advancement , I think we ’ ll see this type of advanced AI-powered automation on the warehouse floor in the near-to-mid-term . The question is how long it will take for the new tech to oust the old - and I think the jury is still out on this .
For example , non-flying drones are increasingly commonplace in warehouses to carry out tasks like stock taking . They ’ re preprogrammed with the details of their environment and use sensors to avoid hazards , and they ’ re often deployed out of hours , making them hugely efficient . 3PLs may see no good reason to turf these machines out and replace them with more expensive AI humanoid robots .
4 . How is AI going to be a gamechanger for the sector ? Generative AI and large language models ( LLMs ) have amazing potential in the field of robotics . AI-based image recognition is coming on in leaps and bounds , and when you put this technology into robots , you ’ re going to get a machine that is far more capable of adapting to its surroundings .
LLMs enable non-technical humans to converse with and issue interpretable commands to machines . Again , this could be transformative on the warehouse floor , allowing existing warehouse teams to direct the robots in the moment rather than requiring engineers to pre-program them beforehand .
10