What we can learn from a logistics giant and a robot startup
(Want to get insights into emerging tech on a more regular basis? Sign up for the official Traction Report newsletter here).
The world of logistics — the commercial activity of transporting goods to customers — where a ‘customer’ can be at any point in a global supply chain — is undergoing enormous change. There are many factors involved, but the two major forces at work are these:
- The continuing rise of e-commerce is leading to logistics firms, like Fedex, UPS, and DHL — and that new kid on the block, Amazon — gearing up their operations to meet that exploding demand. (The corresponding and complementary decline in physical retail sales is the other side of this coin.)
- The technologies that underlie autonomy — like LiDAR sensors, autonomous warehouse robotic systems, and the just-emerging capabilities of autonomous trucks — are revolutionizing the world of logistics.
I will explore the world of logistics by looking at two companies: Fetch Robots, a startup making autonomous mobile robots, and DHL, the world’s largest logistics company. Those end points will shed light on the revolution going on.
Startup of the Week
This week’s startup is Fetch Robotics (fetchrobotics.com), the manufacturer of the Freight series of Autonomous Mobile Robots (AMRs), applicable to material transport and automated data collection. Fetch recently closed a Series B funding round of $25M, led by Sway Ventures with participation from existing investors O’Reilly AlphaTech Ventures, Shasta Ventures and SB Group US (‘Softbank’). Fetch Robotics’ total funding so far is $48 million.
Underlying Fetch’s technology is the FetchCore cloud platform, positioned as an extremely fast solution to get up and running. The company says that at RK Logistics, the AMRs were up and running in a matter of days and now handle 30% to 50% of items in their facility. Fetch customers can also add or modify workflows with a few clicks without programming, and therefore they can adapt in real time to changing conditions.
In the mindmap below, I have categorized Fetch as an example of Mobile Robots, which have more in common with autonomous vehicles than they do with Stationary Robots—which are not shown here, but would be under the Industrial category. This is because their mobility requires the same technology stack as an autonomous car or truck: sensors (LiDAR, vision, etc.), mapping, collision avoidance, and so on.
Note that DHL is a client of Fetch, as are a growing list of other companies, like RK Logistics, and Wärtsilä.
In order to understand the changes going on in logistics in general, we need to zoom in on two sorts of technological disruption that are having an enormous impact on logistics and transportation in general.
Back to the Stack
LiDAR is short for Light Radar, and is a critical technology in the autonomy stack. It’s a remote sensing method that uses pulsed lasers to measure distance, and is key to autonomy. These are like human eyes, that autonomous vehicles use to ‘see’ their surroundings, and from that calculate things like the trajectories of nearby objects and locations of stationary objects like trees next to a roadway or shelves in a warehouse. LiDAR sensors are often located on the roofs of cars, as multi-beam spinning devices, and from their signals and AI-analysis of them, the vehicles create a 3D map of their surroundings.
LiDAR, along with the enormous innovations in AI, are setting the context for a complete overhaul of logistics. As AI is being wed with conventional vehicles, we need to understand the levels of autonomy available today, and how far away we are from driverless trains, airplanes, and ships —>that carry goods great distances —> that are unloaded by robots and placed into autonomous trucks —>and driven to autonomous stores, —> or directly to our homes and workplaces. There’s a ways to go before we see that utopian reality.
Levels of Autonomy
In the world of autonomous vehicles a semi-standard model for levels of autonomy has emerged:
- Level 0: No automation — Also thought of as ‘hands on’. totally steered and monitored by people, with no other options. Basically, your grandma in a 1968 VW Beetle.
- Level 1: Driver assistance required — Think ‘adaptive cruise control’: when you come up on a slow driver in the left lane when adaptive cruise control is engaged, your car will slow down without your involvement. Likewise, ‘lane keeping assistance’ that guides you back into the lane when you drift out of it. But you have to remain completely focused on the car, and remain in control of the vehicle.
- Level 2: Partial automation — Systems at this level will manage speed and steering under specific, limited conditions. They can match speed to cars around you, and follow the curve of the road in good conditions. But you must remain vigilant and prepared to take back full control if the system can’t deal with the situation. This is the level found in Tesla Autopilot, Cadillac Suprecruise, Mercedes Benz Drive Pilot and others.
- Level 3: Conditional Automation — Audi has released the new A8, stating it is the first Level 3 automated vehicle available. The A8 can control your car in a traffic jam, by actively monitoring the situation, but once traffic gets back to over 37 mph, the driver would have to regain direct control of the car. This is the scariest sort of autonomy, because it requires a non-involved passenger possibly becoming the driver again.
- Level 4: High Automation — These vehicles basically drive themselves except in rare conditions, such as extreme wether situations, or contexts that are completely unexpected. While these vehicles still have steering wheels and pedals, these controls may not be used frequently by human drivers.
- Level 5: Full automation — These are fully driverless vehicles that can drive anywhere, anytime. For example, a car coming to pick you up at work after driving from your home, and where you get in the back seat and sleep all the way home.
At present, there are a few level 3 vehicles available, but no level 4 or 5 vehicles, despite marketing hype.
However, to make a serious impact in logistics, there are several intermediate stages that represent stepping stones towards full autonomy, and which could have significant benefits for all involved.
For example, platooning of trucks on highways is already available, where small groups of trucks can form ‘pelotons’ — so that only the foremost driver would be controlling the platoon while other drivers could rest or at least take their hands off the wheel and eyes off the road (the company Peloton is pioneering this approach). This does not line up exactly with the five levels of automation, but is rather a specialized variant of level 1.
Level 4 trucks that can operate autonomously just in the context of highway driving represent a major breakthrough, since they would only require human operators to drive the trucks to and from the highway on surface streets, or directly from distribution sites located immediately adjacent to highways. We can expect this specialized variant of level 4 autonomy before all others.
My recent education into the present and future of autonomous logistics was deepened by a recent event sponsored by DHL, the international logistics leader, where they shared their plans in the market. The following mindmap shows DHL as both a user of autonomous technologies as well as developing their own solutions, like the recently announced Streetscooter, which I describe in more detail, below.
I attended DHL’s press event announcing a new product offering, Parcel Metro, which is an integrated effort by DHL eCommerce to orchestrate an offering based on the surge in e-commerce shipping. Parcel Metro is a DHL service offered in major US cities — currently Chicago and New York City, with other cities to come — that enables same day (Metro Same Day) and next day (Metro Next Day) delivery experience. DHL is also planning to offer an immediate delivery service — Metro Now — in the near future.
The interesting thing about this announcement is that DHL will not be providing the delivery fleet: they will be using other companies — local delivery services and other logistics firms — but will be basically offering their logistics system as a platform for e-commerce or retail organizations to leverage. When I asked if they would ever act as the delivery partner — since the company does deliver packages in the US that are shipped internationally — I was told that DHL is not getting back into US deliveries of domestic goods, after shutting down domestic operations in 2008.
However, DHL is expanding in a number of ways, reflecting its ambitions in a rapidly shifting world.
DHL has an enormous logistics supply chain, relying on hundreds of vendors. They’ve ordered 10 or more Tesla Semis, and are using autonomous warehouse bots from Fetch Robots and Fetch competitor Locus. In time, we can anticipate that every element of their systems will be as automated as possible. At the event, they demonstrated modern robot arms that can pack boxes using techniques that were unavailable until recently, as well as advanced techniques for AMRs to optimally collect packages in distribution sites.
But DHL is also becoming a manufacturer of vehicles: the company has developed and deployed its new Streetscooter vehicles, a modern urban delivery vehicle, designed to be more energy efficient and sustainable. Today’s offerings run on hydrogen fuel cells, and will be transitioning to electric drives in the near term. And the company is partnering with others — like chipmaker Nvidia, to create a world-class autonomous version of Streetscooter as soon as is practical.
DHL is demonstrating that even the most well-positioned and largest logistics companies must actively take on the challenges of autonomous logistics if they are to avoid becoming obsolete in this new era. And they have accepted the challenge and are pushing ahead, as Package Metro and Streetscooter show.
(Want to stay on top of the latest emerging technologies for the enterprise? Join the Traction Tech Council, and stay one step ahead of your competitors).