Compare Stretch, Pickle Unload Robot and Mujin TruckBot by freight, conveyor handoff and dated deployment evidence. Keep advertised rates separate from accepted receiving output.
DROID RESEARCH · SOURCES CHECKED
This comparison covers loose, floor-loaded cases. Moving intact pallets, picking individual items and building outbound orders are different tasks. These three reviewed systems are a starting cohort, not an exhaustive market census.
Grip a case from the load. Record falls, jams and assistance.
Move and identify the case. Record scan failures and blocked transfers.
Deliver to the agreed destination. Count accepted cases, including recovery time.
Droid’s proposed measurement boundary: use the last accepted handoff as the output counter. Log the time and staffing at each step.
Screen a sample case ↗Does an unloading specification also describe the case-picking system?
2025 unloading brochure; current case-picking page tracked separatelyPickle Unload Robot ↗What does the published package envelope leave unresolved?
Current Unload Robot / Dill overview; exact arm and robot revisions unknownMujin TruckBot ↗What has to match when the unloader attaches to a conveyor?
Current TruckBot application page; installed hardware/software revision unknown| System | Documented path | Resolve before ordering |
|---|---|---|
| Stretch | Mobile case handling; current site also describes case picking. Boston Dynamics ↗ | Which application, hardware revision, conveyor and software package are included? |
| Pickle | Mobile unloader with a named KUKA arm and Dill software. Pickle Robot ↗ | Which exact arm, gripper, freight envelope and human-assistance scope apply? |
| TruckBot | Unloader attached to a telescoping conveyor. Mujin ↗ | Which conveyor model, dock geometry, controls and receiving interface have been assessed? |
Each model record retains its supplier’s stated rate and source. These figures use different or undisclosed freight, assistance and timing conditions. They cannot establish which complete receiving line performs best at your site.
For a comparable trial, define the same freight mix and count all scheduled trial time: positioning, trailer changes, interventions, damage checks and downstream stops. Record robot handling time separately so the bottleneck can be located.
These dated records do not establish current fleet size or a service-level guarantee. None is a Droid-run hardware test.
DHL reports commercial North American use from 2023 and subsequent UK/European expansion. It reports up to 700 cases/hour.
Operating mode: autonomous unloading reported; total human workload unspecified.
Evidence limit: This is a customer report, not a matched independent test or a sustained shift average. Exact installed revisions and freight denominator are not provided.
DHL Group ↗The announcement describes an MOU for more than 1,000 additional robots.
Operating mode: future deployment scope.
Evidence limit: An expansion plan is not an installed fleet count. The same announcement describes case picking as an area to explore; the current Boston Dynamics page now describes that application.
DHL Group ↗The suppliers report integration from unloading through conveyor induction, identification, scanning and stacking.
Operating mode: autonomous flow claimed; intervention workload undisclosed.
Evidence limit: No named customer site, observation duration, throughput denominator or independent acceptance report is supplied. This does not establish compatibility with every conveyor or receiving system.
Pickle Robot / Ambi Robotics ↗No named TruckBot operating site is established by the sources in this cohort. That is a coverage gap, not evidence that no deployments exist.
Share your environment, the job to be done, and the outcome you need to measure.