Predictika · Robot cell quotes

An AI quoting agent's robot cell quote, checked before it goes out

An AI quoting agent prepares a robot cell quote for a Yaskawa Motoman GP-series arm: the arm, a controller, the weight of the gripper and part, the reach the cell needs and the discrete inputs and outputs it wires. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of Yaskawa's published figures and returns a verdict, every reason and the figures it compared.

Scope: Yaskawa Motoman GP7, GP8, GP12, GP25 and GP50 arms with the YRC1000 and YRC1000micro controllers. See everything this validator covers

This validator models Yaskawa Motoman GP-series robot cell quotes, not just the scenarios shown below.

  • Arms and controllers — the GP7, GP8, GP12, GP25 and GP50 with the controllers Yaskawa lists for each: the YRC1000 for all five, and the YRC1000micro for the GP7, GP8 and GP12 only.
  • Payload and reach — the published payload (7, 8, 12, 25 and 50 kg) and reach (927, 727, 1,440, 1,730 and 2,061 mm) of each arm against the gripper and part weight and the reach the cell needs.
  • Discrete I/O — the standard points of each controller (40 inputs and 40 outputs on the YRC1000, 8 and 8 on the YRC1000micro); more points are listed as an approval item.

Any GP-series cell quote that touches these rules is in scope — not only the scenarios pictured here. Not covered: other Motoman series (HC collaborative arms, MPX), wrist torque and inertia, options, prices and lead times. The validator shows a cell summary and no line-item table.

The AI quoting agentBuilds the robot cell quote from the conversation and says it is ready.
Predictika validatorChecks every rule at once and returns the verdict, each finding, and the computed figures.

6 scenarios in 3 groups · each group runs from the agent’s first attempt to the outcome · click one to switch

What the agent proposes

The AI quoting agent says

Illustrative agent statement. The validator receives the structured request below.

Proposing agent: AI quoting agent (scripted)

Before sending, the agent must call Recorded run

Predictika validator


      

        
        
        
      

      
Waiting for the call.

      
    

  

  
  

    
Where rules-based CPQ gets it wrong
Click to expand
    
Our assessment of what a typical rules-based CPQ (product rules, price rules and validation rules, as in Salesforce CPQ) does with the same cases. It is a judgment about effort and reliability, not a claim that nothing can be built.

    
What the validator caughtTypical rules-based CPQPredictika
Controllers that are listed for only some armsOnly with care
Each arm needs a rule naming the controllers it may use, and the list has to be rechecked whenever an arm or a controller is added.		The arm and the controller are in one request, and the listed pairs are checked together on every call.
Payload and reach limits that depend on the armOnly with care
Each arm carries its own limits, so every arm needs its own payload rule and its own reach rule.		The arm is a field of the request, and its published payload and reach are compared with the cell on every call.
A fix that exposes the next problemOnly with workarounds
Changing the arm changes which controllers and limits apply, and rules fire as each choice is made, so the next error appears only after the first is fixed.		The whole cell is rechecked from the request on every call, so a new arm shows its controller problem at once.
Hard limits next to limits that need approvalOnly with workarounds
Blocks and approvals need separate rule types and a status that combines them, so the answer is easily wrong.		Block reasons and approval items come back together, with one verdict.
Every problem returned in one answerOnly with workarounds
Rules tend to fire as the quote is saved or a step is completed, so a rep often fixes one error and meets the next.		All findings come back together, whatever order the choices arrived in.

  

  

    
What this validator catches: findings from its test cases
Click to expand
    
Every rule is checked on every call, in any order the choices arrive; one problem never hides another. Blocking findings stop the quote; approval findings need a person to sign off; everything else is computed.

    
1The gripper and part weigh 8 kg but the GP7 is rated for 7 kg.
2The cell needs 728 mm of reach but the GP8 reaches 727 mm.
3The cell needs 1,441 mm of reach but the GP12 reaches 1,440 mm.
4Yaskawa lists the GP25 for use with the YRC1000 only, not the YRC1000micro.
5Yaskawa lists the GP50 for use with the YRC1000 only, not the YRC1000micro.
6The gripper and part weigh 51 kg but the GP50 is rated for 50 kg.
7The cell needs 2,062 mm of reach but the GP50 reaches 2,061 mm.
8The cell needs 9 inputs and 8 outputs but the YRC1000micro provides 8 and 8 as standard; the extra points need an additional I/O board or a fieldbus board.
9The cell needs 41 inputs and 40 outputs but the YRC1000 provides 40 and 40 as standard; the extra points need an additional I/O board or a fieldbus board.
10The gripper and part weigh 55 kg but the GP50 is rated for 50 kg.
11The cell needs 2,100 mm of reach but the GP50 reaches 2,061 mm.
12The cell needs 20 inputs and 20 outputs but the YRC1000micro provides 8 and 8 as standard; the extra points need an additional I/O board or a fieldbus board.
13The cell needs 800 mm of reach but the GP8 reaches 727 mm.
14The cell needs 12 inputs and 12 outputs but the YRC1000micro provides 8 and 8 as standard; the extra points need an additional I/O board or a fieldbus board.
15The gripper and part weigh 9 kg but the GP7 is rated for 7 kg.
16The cell needs 1,000 mm of reach but the GP7 reaches 927 mm.
17The cell needs 12 inputs and 6 outputs but the YRC1000micro provides 8 and 8 as standard; the extra points need an additional I/O board or a fieldbus board.

  

  

    
Scenarios and results
Click to expand
    
The recorded result of the validator for each scenario on this page: the request the AI agent sent, the verdict, and every finding, in the order they appear above.

    
1. Arm, controller and I/O limits: A GP7 cell with a 20 kg part, 1,500 mm of reach and 12 inputs and outputs on the compact controller

    
The agent's request. Please quote a robot cell for Northline Packaging: a GP7 arm on the Motoman YRC1000micro controller. The gripper and part weigh 20 kg together, the arm has to reach 1,500 mm, and the cell wires 12 discrete inputs and 12 discrete outputs. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Payload too high. The gripper and part weigh 20 kg but the GP7 is rated for 7 kg.
    • 
      
  • Reach too short. The cell needs 1,500 mm of reach but the GP7 reaches 927 mm.
    • 
      
  • More I/O than the controller provides. The cell needs 12 inputs and 12 outputs but the YRC1000micro provides 8 and 8 as standard; the extra points need an additional I/O board or a fieldbus board.
    • 
    

    
2. Arm, controller and I/O limits: The agent resubmits after fixing the errors

    
The agent's request. Please quote a robot cell for Northline Packaging: a GP25 arm on the Motoman YRC1000micro controller. The gripper and part weigh 20 kg together, the arm has to reach 1,500 mm, and the cell wires 12 discrete inputs and 12 discrete outputs. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Controller not listed for this arm. Yaskawa lists the GP25 for use with the YRC1000 only, not the YRC1000micro.
    • 
      
  • More I/O than the controller provides. The cell needs 12 inputs and 12 outputs but the YRC1000micro provides 8 and 8 as standard; the extra points need an additional I/O board or a fieldbus board.
    • 
    

    
3. Arm, controller and I/O limits: The agent resubmits after fixing the errors

    
The agent's request. Please quote a robot cell for Northline Packaging: a GP25 arm on the Motoman YRC1000 controller. The gripper and part weigh 20 kg together, the arm has to reach 1,500 mm, and the cell wires 12 discrete inputs and 12 discrete outputs. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Cell summary. GP25 on the YRC1000: 20 kg part against 25 kg rated, 1,500 mm of reach needed against 1,730 mm, and 12 inputs and 12 outputs needed against 40 and 40 as standard.
    • 
    

    
4. Reach: A GP8 cell needing 800 mm of reach

    
The agent's request. Please quote a robot cell for Northline Packaging: a GP8 arm on the Motoman YRC1000micro controller. The gripper and part weigh 8 kg together, the arm has to reach 800 mm, and the cell wires 4 discrete inputs and 4 discrete outputs. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Reach too short. The cell needs 800 mm of reach but the GP8 reaches 727 mm.
    • 
    

    
5. Reach: The agent resubmits after fixing the errors

    
The agent's request. Please quote a robot cell for Northline Packaging: a GP12 arm on the Motoman YRC1000micro controller. The gripper and part weigh 8 kg together, the arm has to reach 800 mm, and the cell wires 4 discrete inputs and 4 discrete outputs. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Cell summary. GP12 on the YRC1000micro: 8 kg part against 12 kg rated, 800 mm of reach needed against 1,440 mm, and 4 inputs and 4 outputs needed against 8 and 8 as standard.
    • 
    

    
6. I/O beyond the standard points: A GP12 cell on the compact controller needing 12 inputs and 6 outputs

    
The agent's request. Please quote a robot cell for Northline Packaging: a GP12 arm on the Motoman YRC1000micro controller. The gripper and part weigh 9 kg together, the arm has to reach 1,000 mm, and the cell wires 12 discrete inputs and 6 discrete outputs. It is ready to send.

    
Verdict: Needs approval.

    
    
      
  • More I/O than the controller provides. The cell needs 12 inputs and 6 outputs but the YRC1000micro provides 8 and 8 as standard; the extra points need an additional I/O board or a fieldbus board.
    • 
      
  • Cell summary. GP12 on the YRC1000micro: 9 kg part against 12 kg rated, 1,000 mm of reach needed against 1,440 mm, and 12 inputs and 6 outputs needed against 8 and 8 as standard.
    • 
    

  

  

    
Narration transcript
Click to expand
    
The text of the narrated walkthrough, step by step.

    
    
      
  1. Welcome to a demo of Predictika's Validator. An AI quoting agent prepares a robot cell quote for a Yaskawa Motoman GP-series arm: the arm, a controller, the weight of the gripper and part, and more. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of Yaskawa's published figures and returns a verdict, every reason and the figures it compared.
    2. 
      
  2. I will start with the first request by the agent.
    3. 
      
  3. The agent says. Please quote a robot cell for Northline Packaging: a GP7 arm on the Motoman YRC1000micro controller. The gripper and part weigh 20 kg together, the arm has to reach 1,500 mm, and the cell wires 12 discrete inputs and 12 discrete outputs. The rest of the request is on screen.
    4. 
      
  4. Before it goes out, the agent calls the validator.
    5. 
      
  5. Blocked, with two findings and one approval, all in one answer. The gripper and part weigh 20 kg but the GP7 is rated for 7 kg. The cell needs 1,500 mm of reach but the GP7 reaches 927 mm.
    6. 
      
  6. The validator returned two blocking findings and one approval in one answer. A check that reads one field at a time would pass most of this; it shows only when the choices in the quote are checked against each other.
    7. 
      
  7. What this shows: for payload and reach limits that depend on the arm, a rules-based CPQ is error-prone for this.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the robot arm to GP25.
    10. 
      
  10. It calls the validator again.
    11. 
      
  11. Still blocked, with one finding and one approval, all in one answer. Yaskawa lists the GP25 for use with the YRC1000 only, not the YRC1000micro.
    12. 
      
  12. Two findings cleared, but the change brings a different blocking finding: controller not listed for this arm.
    13. 
      
  13. The agent resubmits once more.
    14. 
      
  14. The revised request. It changes the controller to Motoman YRC1000.
    15. 
      
  15. It calls the validator once more.
    16. 
      
  16. This time the validator clears it.
    17. 
      
  17. Both earlier findings are cleared.
    18. 
      
  18. Three problems caught in one call, fixed in two passes, and a clean quote. That is the value of Predictika's validator: catching mistakes before they reach the customer. Now you can try the other scenarios yourself.
    19.