Tao IP
Patentability Report: a warehouse robotics startup
Report score
A good candidate
The strongest candidate's composite score: how much of it is still open against the closest references found. Higher means more room.
0 early20 worth a disclosure40 a good candidate60 a strong candidate80 a priority
By candidate
- Candidate 158%
- Candidate 234%
- Candidate 332%
Summary
Three Candidates were evaluated:
- Candidate 1 (grasp selection for mixed totes) scores highest at 58%. The closest patent scores grasp points for one fixed arm. Choosing among suction, pinch and push for each item, and changing the approach after a miss, is only partly shown.
- Candidate 2 (slotting from pick history) scores 34%. Slotting art is crowded, and claims on the scheduling logic alone take a weaker §101 position than claims tied to the robots.
- Candidate 3 (damage checks on outbound cartons) scores 32%. A parcel carrier's patent already shows camera inspection at the dock door.
| Direction | Closest reference | Novelty* | Eligibility** | Composite |
|---|---|---|---|---|
| 1. Grasp selection and recovery for mixed totesUS patent, 2019, a large robotics company | US patent, 2019, a large robotics company | 0.68 | 0.85 | 58% |
| 2. Slotting and robot routing from pick historyUS application, 2021, a logistics software company | US application, 2021, a logistics software company | 0.55 | 0.62 | 34% |
| 3. Damage checks on outbound cartonsUS patent, 2017, a parcel carrier | US patent, 2017, a parcel carrier | 0.40 | 0.80 | 32% |
*Novelty is an estimate of how likely a claim of commercially useful scope is to survive §102/§103 analysis over the references found.
**Eligibility is an estimate of how likely a claim is to survive §101 analysis.
Public record
The company builds a picking cell for mixed totes: a mobile base, a gripper with suction and pinch, and software that plans each pick. It has posted a demo video and two customer case studies, and a founder spoke about grasp recovery at a robotics conference this spring.
What the company already holds
One published US application from 2023 on tote presentation hardware, with both founders as named inventors. We treat it as the founders' own work and the starting point a new filing has to go beyond, never as a competitor's.
Why file now
- The demo video went public in June. Anything shown in public before a filing counts as prior art against that filing.
- A large ecommerce retailer published two applications on mixed tote picking this year, close to Candidate 1.
Disclosure
| Direction | Candidate invention | Features compared |
|---|---|---|
| 1 | Choosing suction, pinch or push for each item in a mixed tote, with recovery after a failed pick | F1.1 Grasp mode scoring for each itemF1.2 A push that separates items before a pickF1.3 Recovery that changes the approach after a missF1.4 Learning from pick outcomes across sites |
| 2 | Slotting inventory and routing robots from the same pick history | F2.1 Slot moves ranked by pick frequencyF2.2 Robot routes planned against the new slotsF2.3 Moves timed to idle robots |
| 3 | Checking outbound cartons for damage before they leave the dock | F3.1 Images from the pick cell reused at packingF3.2 A damage score for each cartonF3.3 A hold and repack rule |
Top 10 closest references
Ranked by how many features across the candidates a reference shows. Each reference links to the public record.
| Rank | Reference | Owner | Priority | What it shows | Directions |
|---|---|---|---|---|---|
| 1 | US patent: grasp planning for bins of mixed items | A large robotics company | 2019 | Scores grasp points for suction and pinch from a depth image, then retries at the next best point. A fixed arm at a station, and it never pushes items apart before a pick. | 1, 3 |
| 2 | US application: tote presentation for a picking robot | The company (own filing) | 2023 | The tote tilt and camera mount the cell uses today, with both founders as named inventors. The starting point new filings have to go beyond. | 1 |
| 3 | US application: slotting from order history | A logistics software company | 2021 | Moves fast selling items closer to the pick stations from order history. Planning robot routes against the new slots is not described. | 2 |
Ranks 4 to 10 follow in your report, each linked to the public record.
Candidate 1: grasp selection for mixed totes (58%)
The closest patent scores grasp points for one fixed arm and retries at the next best point. The company's own application covers the tote hardware. What remains open is scoring suction, pinch and push for each item on a mobile base, pushing items apart before a pick, and changing the approach after a miss.
Shown by the closest reference: yes, partly or no. No leaves the most room.YesPartlyNo
| Feature | Closest reference | Shown? | What the reference shows / gap |
|---|---|---|---|
| F1.1 Grasp mode scoring for each item: suction, pinch or push, from shape and surface cues | US patent, 2019 | Partly | Scores grasp points for suction and pinch. Push as a mode, chosen item by item, is not shown. |
| F1.2 A push that separates items before a pick | US application, 2024 | Partly | Shakes the tote to separate items. A planned push by the gripper is not described. |
| F1.3 Recovery that changes the approach heading and gripper after a miss | US patent, 2019 | No | Retries at the next best point with the same approach. Changing heading or gripper is not shown. |
| F1.4 Learning from pick outcomes across sites | US patent, 2016 | Yes | Logs pick outcomes from many cells and updates one shared grasp model. |
What would raise the score: Describe the scoring function and its inputs, what changes after a miss, and the measured gain in first pick success. Claim the mode set this gripper actually has.
Your market on the patent record
6 companies filed 214 US applications in the company's technology classes since 2024. We found 1 from the company.
Who files in your classes
| Company | 2024 | 2025 | 2026 | Total |
|---|---|---|---|---|
| A large ecommerce retailer | 31 | 38 | 27 | 96 |
| A large robotics company | 18 | 16 | 12 | 46 |
| A warehouse automation company | 9 | 12 | 10 | 31 |
| A robot gripper maker | 6 | 8 | 7 | 21 |
| A parcel carrier | 4 | 5 | 3 | 12 |
| A logistics software startupNew | 0 | 2 | 6 | 8 |
| This company | 0 | 1 | 0 | 1 |
New entrants
Companies that started filing in these classes in the last 18 months.
- A logistics software startup, 8 recent publications
- A university spinout in robotic grasping, 3 recent publications
Tao IP Watch follows these companies every week.
Scope and method
Tao AI read what the company has published and its own filings, then searched published patents and applications for each candidate and read the closest ones in full. Nothing that is not public was used.
Sources
- The company's website, product page and demo video
- USPTO records for the company and its founders
- The ten closest references, read in full
- USPTO application counts in the company's technology classes
Prepared by Tao AI at Tao IP from public information. Software output, not legal advice.
