Whereas a lot of the present dialogue round bodily AI focuses on applied sciences nonetheless transferring from prototype to business deployment, AES has already put robots to work on utility-scale photo voltaic development websites.
The worldwide vitality firm developed Maximo, an AI-powered robotic system designed to automate one of the crucial bodily demanding levels of photo voltaic development: lifting and putting in photovoltaic modules.
Quite than working within the predictable atmosphere of a manufacturing facility, Maximo works alongside development crews on energetic websites, the place terrain, climate, gear and dealing circumstances can change consistently.
The system has now put in greater than 180,000 photo voltaic modules representing greater than 100 MW of producing capability. In keeping with AES, robots utilizing Maximo have exceeded 500 installations per day, whereas crews utilizing the know-how can set up as much as 24 modules per particular person per shift hour.
Maximo was developed via AES Subsequent, the corporate’s innovation and business-building platform. The undertaking gave its builders an uncommon benefit for a robotics enterprise: entry to AES’s personal pipeline of reside photo voltaic development initiatives, permitting the know-how to be examined and refined below actual working circumstances from an early stage.
Nick Hegeman, chief business officer for Maximo, is answerable for serving to take that know-how from inner growth and area validation towards broader business adoption.
On this Q&A with Robotics & Automation Information, Hegeman discusses the technical challenges of deploying autonomous methods on altering development websites, why area validation is important to the success of bodily AI, and the way labor shortages and more and more heavy photo voltaic modules are strengthening the case for automation.
He additionally explains why integration with current development processes has been central to Maximo’s growth, how prospects’ attitudes towards robotics have modified as deployments have scaled, and the place AES sees the following alternatives for automation throughout renewable vitality development and website logistics.
Interview with Nick Hegeman
Robotics & Automation Information: Building has lengthy been considered one of the crucial troublesome industries to automate as a result of each website is completely different. What had been the most important technical challenges Maximo needed to overcome earlier than it may reliably deploy robots on reside photo voltaic development initiatives?
Nick Hegeman: Photo voltaic development operates below extraordinarily strenuous circumstances, in open desert and infrequently in excessive warmth. On the Bellefield undertaking, Maximo helped crews set up 180,000 particular person modules, with each weighing as much as 75 kilos. Crews must elevate that a whole lot of instances a day.
The problem was constructing one thing dependable sufficient to try this job at that tempo with out asking EPCs to vary something about their processes.
On prime of that was the sheer selection throughout websites, completely different terrain, completely different climate, completely different module fashions and wattages from one undertaking to the following, so the system couldn’t depend on being reprogrammed or retrained each time it confirmed up someplace new.
It wanted to take a look at no matter was in entrance of it and know the depth, tilt, and placement by itself. We didn’t wish to change the provision chain or reinvent the wheel, we needed Maximo to fit into current development workflows so adoption can be as easy as attainable.
R&AN: Your robots have now accomplished greater than 180,000 photo voltaic panel installations and surpassed 100 MW of autonomous deployment. At what level did prospects cease viewing the know-how as an experiment and start treating it as a sensible development software?
NH: Maximo was incubated as a part of AES Subsequent. This has supplied our group with the chance to deploy pilot packages at varied photo voltaic initiatives throughout the US. Once we moved from pilots to utility-scale initiatives, that shift turned obvious. Prospects are actually relying on us to fulfill deadlines.
In December 2025 and January 2026, we reached an set up tempo of 20MW a month. Two of the 5 largest EPCs within the nation got here out to see Maximo operating at this tempo, and in each circumstances, they instantly fast-tracked their know-how analysis course of.
Watching the robots carry out at scale in actual circumstances moved Maximo from “price a glance” to “price prioritizing” of their eyes. In the present day crews utilizing Maximo are putting in as much as 24 modules per particular person per shift hour, and EPCs are constructing that efficiency functionality into their customary development workflow.
R&AN: In contrast to autonomous autos working on mapped roads, photo voltaic development websites are consistently altering, with uneven terrain, transferring gear and folks working close by. How does Maximo allow its robots to function safely and effectively in such unpredictable environments?
NH: Security with Maximo begins earlier than the robotic is ever switched on. Crews undergo coaching and site-specific security checks as a part of customary setup, the identical manner any new piece of heavy gear can be launched to a job website.
As soon as it’s working, Maximo layers real-time notion on prime of that basis. Max repeatedly identifies the tracker construction, modules, and something close by, and self-corrects the second circumstances change, whether or not that’s an individual coming into the work space or an sudden obstruction.
R&AN: Labor shortages stay a serious problem throughout each the development and renewable vitality sectors. Has your expertise proven that prospects are adopting robotics primarily to scale back labor prices, to deal with workforce shortages, or to extend undertaking supply velocity?
NH: All of these things are associated, however the important thing EPC’s are specializing in is preserving their contractual commitments to ship a undertaking on time for a predetermined value. Their primary job is on the lookout for methods to derisk a undertaking, however the realities of the development atmosphere make this more and more troublesome.
All prospects need the advantages of robotics, whether or not it’s velocity, prices or security, however till not too long ago there’s at all times been the query as to what dangers does introducing robotics carry to the undertaking. What’s actually accelerated buyer adoption for Maximo is the benefit of integration and dependable efficiency.
These initiatives are already operating so tight that the majority EPC’s can’t afford to take a danger on one thing that’s unproven. Our observe file of efficiently putting in a whole lot of hundreds of modules is what takes robotics from one thing that’s simply one other factor PM’s should fear about to one thing that’s making their lives simpler.
Prospects don’t select robotics for a single purpose anymore, speed-to-power, labor availability, and panel weight have all converged into one downside, and proper now, speed-to-power is the priority sitting on prime of every thing else.
Maximo addresses that stress by delivering set up charges greater than double the trade common, reduces the crew dimension wanted for heavy modules, and may shave as much as two months off utility-scale timelines.
Whereas Maximo can not exchange the necessity for employees, it may give crews a method to shut a part of that hole and transfer quicker with the labor they have already got.
R&AN: Many robotics firms are actually selling “bodily AI”, however comparatively few have reached business deployment at scale. Out of your perspective, what separates applied sciences that succeed within the area from people who stay spectacular demonstrations?
NH: Loads of firms on this area have raised tens of millions in funding and nonetheless haven’t gotten a robotic onto a reside job website. The applied sciences that make it previous the demo stage are those constructed with the sphere in thoughts from day one, which suggests rigorous testing nicely earlier than a robotic encounters a reside website.
Simply as necessary is buy-in, and that’s the place being constructed inside AES Subsequent made a distinction. Maximo went via AES’s personal structured funding and incubation course of, the identical mannequin that helped construct different companies like Fluence.
Maximo was designed to fit into current development workflows and provide chains fairly than drive EPCs to vary how they construct, which is why adoption has been easy fairly than a tough promote. Bodily AI succeeds commercially when the individuals utilizing it belief it, and that belief is earned via testing, coaching, and security self-discipline.
R&AN: Maximo has demonstrated how robotics can automate one stage of utility-scale photo voltaic development. Wanting forward, which different elements of renewable vitality infrastructure do you imagine are prepared for automation, and the place do you see the best alternatives for robotics over the following decade?
NH: Photo voltaic module set up was the primary bodily stage prepared for automation, but it surely isn’t the one one. We’re additionally seeing automation take maintain on the digital aspect of those initiatives, giving EPCs higher real-time visibility into set up progress and crew efficiency as work occurs, not simply after the actual fact. That’s a separate piece of the story we’ll have extra to share on quickly.
Wanting additional forward, materials dealing with and website logistics stand out as the following main alternative. That covers every thing concerned in getting the bodily elements the place they have to be, delivering modules and pylons, receiving the elements required for set up, and transferring all of it round an energetic development website.
It’s a much less seen a part of the method when in comparison with set up but it surely’s simply as important, and the know-how behind it’s actually near being automated at utility-scale.
We’re additionally preserving a detailed eye on modular and tracker OEMs. Panels proceed to get bigger and heavier. A number of manufactures are already beginning to modify their designs with automated set up in thoughts, which means that robotics is beginning to affect how {hardware} will get designed upstream.
R&AN: AES operates one of many world’s largest vitality companies whereas additionally creating robotics via Maximo. Has being a part of a serious vitality firm modified the best way you develop and commercialize robotics in contrast with a traditional robotics startup, and what classes does that supply the broader trade?
NH: Being incubated inside AES gave us an actual development pipeline to study from, which meant we didn’t must spend years chasing a primary website the best way a standalone startup would.
Maximo was conceived in Virginia, inbuilt Pittsburgh, and put onto AES’ personal energetic photo voltaic initiatives nearly instantly, so as an alternative of testing in a managed atmosphere and hoping to land a pilot, we had been field-proven from day one as a result of we had been engaged on AES’s reside development websites.
That entry lets us prioritize getting Maximo into the sphere as early as attainable fairly than ready for a completed product. We may put it via actual desert warmth, actual terrain, and actual crews straight away and study quick from precise circumstances as an alternative of simulations.
That velocity to the sphere, made attainable by working inside an organization that already had the initiatives and the pipeline, was an actual benefit and takeaway.
