Rather than backing the most eye-catching machine, corporate investors should be looking for technologies that lift throughput, fit existing operations and earn their way across a network.

Robotics in logistics is often presented as a contest of spectacle: ever more capable humanoids, autonomous vehicles and machines that seem to anticipate the future. But the investment case made by Will O’Donnell, managing partner of Prologis Ventures, and JP Keating, founder and managing partner of Prochain Ventures, is more prosaic — and potentially more useful. The form-factor of the hardware is irrelevant — it matters what the system allows the logistics company to do.
“The robot is not the product. The productivity improvement is the product,” said Keating, speaking at a recent GCV webinar on the logistics robot investment opportunities.
A machine can be technically impressive and still fail to improve throughput, reduce errors or lower operating costs. Worse, automating one step without considering the rest of the process can simply move the bottleneck. Faster picking, for example, is of limited value if goods then pile up waiting to be loaded on to lorries.
“If I bring in a robot to increase my picking and packing, but I don’t then address the fact that I need more throughput going out the door — so loading trucks after it’s done — I suddenly have a log jam at the warehouse that it starts breaking,” said O’Donnell.
Keating founded ProChain Ventures after a quarter-century working on the complex supply chains of companies like Boeing and Anheuser-Busch, and now invests in logistics and supply chain technology to solve known pain points for large corporations. O’Donnell, meanwhile, has spent more than 20 years at Prologis, the world’s largest logistics real estate company, building a venture investment arm that uses the company’s billion‑square‑foot network as a testbed for co‑innovation and incubation.
They joined GCV’s webinar: Robotics meets logistics: the frontier for investors to discuss their insights on the current state of the market.

The less visible investment opportunities
Though humanoid robots have captured many of the headlines and largest recent funding rounds in robotics, AI has created a host of less visible but highly significant advances for logistics-focused robots.
One promising area is modular automation. Traditional fixed systems, such as conveyor installations, can be difficult to adapt when a customer’s needs or site change.
“The last four years, we’ve seen a lot of advancements in modularity being driven by AI, and that’s important because then a robot or an automation can be adopted for multiple purposes. So I can take a locusrobot, for example, and pull it out of one warehouse and put it in another,” said O’Donnell.
More flexible systems can be moved or repurposed, giving the equipment a longer useful life and a potential resale value. That can improve the economics for operators and reduce the risk of investing in hardware that becomes stranded.
AI has also made it easier and cheaper to test robotics systems. AI-enabled simulation can let companies run more iterations before deploying equipment on site.
“It’s faster to get and test and pilot and do visualisations. So you can run simulations that before I had to have a robot in a real world to actually test,” said O’Donnell.
While investing in hardware remains expensive, improved testing is helping de-risk the projects and is shortening sales cycles.
Another area of growing interest for investors is operational visibility. Keating described computer vision tools that map movement in warehouses and yards.
“They analyse the entire warehouse, the operations within the warehouse, within a yard. They can see where all the bottlenecks are, what assets aren’t utilised properly, and you don’t need to spend a year doing a human observation to do this,” said Keating.
This kind of technology may help operators decide where automation is genuinely needed — and where a process change could deliver more value than another machine.

Our annual industry survey tracks metrics such as fund and team size, salary and performance across the corporate venture sector globally.
Add your data and receive a free copy of the resulting benchmarking report. All data is anonymised.
A key area of interest is the connective tissue between assets and systems. Warehouses may contain machines from several automation suppliers alongside existing warehouse, transport and yard management software. The investment opportunity includes tools that help those systems work together and show managers the consequences of changes upstream and downstream.
“Supply chains have changed. If you adjust one link, it has upstream and downstream impacts, and it’s really hard to understand what that is until you actually make that change. Back in COVID, there weren’t enough chassis at the port of LA, and suddenly the entire supply chain broke,” said O’Donnell. That kind of bottleneck happens on a smaller scale in warehouses on a daily basis, he said.
“As a manager, I need to be able to make decisions because we’re doing a sale this week, or 10% of my workforce is sick, forklift broke. So how do I really have an orchestration layer that allows me?”
There are also opportunities in making existing infrastructure more productive. Keating pointed to autonomous rail cars that could make underused freight assets more flexible, and to a company developing ways to retrofit existing racking rather than replace it with a proprietary system.
“Instead of having to buy a proprietary racking system, so now you’re talking like a capital outlay of like $5m instead of $100m, and that gets people’s attention,” said Keating.
Margins are razor-thin in logistics and warehousing, which means that not only is the capital outlay of a new robotics a factor, but it is also important to get it up and running quickly.
“If I’m putting in $40m into an automation system, the ROI is pretty long and massive in order to get it. So the payback period may take me five years to pay back. You see another automation, where they can lower the cost of implementation. It’s much more plug and play. You’ve seen some robotics as a service model, which are all intended to take that friction up front to get people to get an ROI faster,” said O’Donnell.
A quick implementation may trump other features, he said.
“If it takes me six months to integrate a robot into my IT system, good luck running a pilot.”
Cobots are practical; humanoids remain conditional
Collaborative robots, or cobots, are already a practical part of the automation conversation. They can take on repetitive, physically demanding or hazardous tasks, while allowing workers to focus on decisions and activities where human judgement is more valuable. The case is not necessarily to remove people from a process, but to reduce the strain and waste within it.
Humanoid robots have a different proposition: versatility. A human-shaped machine could, in principle, work across tasks and environments designed around people. But the form alone does not make it useful. In a flat warehouse, wheels may be more effective than legs; dexterous gripping remains a difficult technical and economic challenge. The strongest cases may be environments where a human form is genuinely useful — for instance, where equipment must navigate obstacles or operate in spaces built for people.
For now, the investment test is task-specific. A specialised machine may deliver a clearer return than a humanoid that promises broader capabilities but has yet to demonstrate reliable performance at an acceptable cost.
“When it comes down to heavy industry, high throughput, high reliability, these are a tough sell at the moment,,” said Keating. “I’m not going to discount them, but it just is going to come down to what that what that job is that has to be solved.”
Six takeaways for corporate investors
- Invest in outcomes, not form factors. Start with measurable gains in throughput, accuracy, safety or cost.
- Assess the whole workflow. A local efficiency can create a downstream bottleneck if the wider process is not considered.
- Look for flexibility and residual value. Modular, mobile and repurposable systems may reduce the risk of stranded capital.
- Back integration and visibility. Simulation, computer vision and orchestration can help make hardware easier to deploy and operate.
- Demand evidence of scale-out. A successful pilot is not the same as adoption across multiple sites, customers and operating models.
- Treat humanoids as a selective bet, not a default. Cobots and purpose-built machines may offer more immediate returns where the task is clear and repeatable.
The central investment thesis is not that every warehouse needs more robots. It is that logistics operators need systems that can adapt, connect and keep goods moving with less friction. In a sector where margins are tight and disruptions are inevitable, the most valuable innovation may be the least theatrical: technology that fits the infrastructure already in place, proves its worth quickly and can be replicated across the network.
Watch the full webinar:
This webinar was part of our monthly series of The Next Wave webinars, exploring corporate venture and innovation. Our next webinar on CVC practice will be in November.
