Two former Carnegie Mellon faculty members, who happen to be brothers, discuss how they are taking a pragmatic approach of adding automation to long-haul trucking. Their company, Locomation, upgrades an existing big rig by replacing the side-view mirrors with a module that includes a mirror, lidar, radar, camera sensors, and radios.
This suite of sensors feeds a computer running their software stack. Their system and service allow trucking companies to field a two-truck platoon. Their initial target market is time-sensitive, long-haul applications where trucking companies normally employ two drivers for a single truck with minimal stops (one sleeps while the other drives).
Improved Efficiency & Safety #
Locomation’s solution allows one driver per truck. One truck and driver lead, while the other driver rests in her truck. At a shift change, the lead vehicle and trailing vehicle swap. This allows trucking companies to effectively double their payload per driver compared to the existing mode of operation, while still staying within the bounds of the FMSCA regulations.
A secondary benefit is the estimated 8% reduction in fuel use, thanks to the reduced drag due to the proximity of the two vehicles.
Improved safety may even more important than economic benefits. Tekin Mericli, Locomation’s CTO, describes what he calls advanced, Advanced Driver Assist System. He says their ADAS combines the strengths of human intelligence with the precision of computer control. Their computers control the brakes, accelerator, and steering wheel of the trailing big rig. It is like a wireless leash connects the two vehicles.
Locomation differs from fully autonomous solutions in that they rely on the lead driver and her experience to be the brains behind the wheel; the cognitive filter as Tekin states. At the same time, the driver is also effectively training Locomation’s software on how to drive, providing a path to full automation. In the not too distant future, the Mericli brothers envision their computers logging millions of miles of real-world driving. In turn, this data will be used to train Locomation’s AI.
The Business Model & Target Market #
Today they are laser-focused on a niche that can benefit from a huge improvement in efficiency without requiring full automation. According to Locomation CEO, Cetin Mericli, their target equals approximately 250,000 out of a market-size of approximately 4 million trucks.
They are looking at a subscription model, whereby they would outfit operators’ existing trucks with their sensors, computers, and software. This minimizes the upfront financial cost to the trucking company while providing Locomation with ongoing cash flow.
Their subscription term is three years, which equals the expected useful life of a truck operating at 90% utilization. Utilization is a big part of the value as Cetin suggests trucks typically are in use only about 30% of the time.
Their recent announcement with Wilson Logistics to begin an Autonomous Relay Convoy pilot this spring is a proof-point of the traction they are gaining. The improved efficiency and reduction in CO2 emissions are great promised benefits, but the best benefit may be the improvement of the work environment for the truck drivers.
[Added 3/30/20 – Question/Concern from a Reader & Locomation Response #
A concern posed by a retired 1+million mile trucker is about the space between the trucks. Specifically, what prevents drivers from trying to insert their cars in that space.? He said that when he pulled double sets of trailers, there were multiple attempts by drivers to cut in between the two trailers which were connected by a tow bar. He cannot imagine how this would be prevented if there wasn’t a tow bar.
In a nutshell, he suggests that any safe spacing between the two trucks would be too tempting and inevitably, some irresponsible driver would try to fill the space.
Even though our current test videos show the gap between trucks to be slightly large, in the production system, the autonomous follower truck will be following the leader from ~20 ft, similar to what’s shown in the concept animation on our website (https://locomation.ai) as well as the interview video. Given that the average length of a passenger car is 15 ft, that tight gap between our trucks should highly discourage people from attempting to cut-in. Also, our vehicles are going to be carrying DOT-issued truck convoy stickers and other signage to warn the surrounding traffic, and it will be illegal to perform maneuvers, such as cut-ins, that would interfere with the operation of those autonomous convoys. Additionally, our sensors will be recording everything that’s happening around the vehicles, and we will have evidence on any such adversarial attack on our vehicles. Safety is of utmost importance for us and we are doing everything we can to keep our drivers, vehicles, and the surrounding traffic as safe as possible.
Highlights of the Above Interview:
- 00:19 – Locomation background and overview
- 02:23 – Autonomous Relay Convoy and human at the center
- 04:20 – How do the safety rules apply?
- 04:58 – Platooning efficiency
- 05:28 – An extremely narrow operational design domain
- 07:06 – Combines the strengths of human and machine
- 08:36 – Modular solution – replace side-view mirrors & add electronics
- 09:50 – The business model and target customers
- 10:52 – Target market and the potential number of miles driven
- 12:04 – A subscription model
- 13:41 – The development schedule – starting in late 2018
- 15:24 – Average team member has 14 years robotics/av experience
- 16:00 – Commercialization from mid-to-late 2021
- 17:43 – Federal versus State regulations
- 18:59 – Not relying on DSRC for their Vehicle-to-Vehicle communications
- 21:27 – Side-view mirror redundancy