Predictika · mtu Series 4000 diesel genset plants for data centers: sizing, emissions, fuel and terms

An AI sales agent's genset plant quote, checked before it goes out

An AI sales agent prepares a quote for a diesel genset power plant for a fictional data center, in euros. Before it goes out, the agent sends it to Predictika’s validator, which checks it against the sizing, emissions, fuel, installation and pricing rules, compares the quote total the agent states with the one it computes, and returns a verdict with every reason.

Scope: mtu Series 4000 diesel genset plant quotes for data centers, in the style of Rolls-Royce Power Systems: sizing, redundancy, emissions, fuel, enclosure and commercial terms. See everything this validator covers

This validator models mtu Series 4000 diesel genset plants sized for data centers. It handles any combination of rating class, redundancy scheme, region, enclosure and service terms that the five modeled genset models can produce, not just the scenarios shown below.

  • Genset sizing and redundancy — Rated kVA by ISO 8528 class for five sample mtu Series 4000 models, facility load from IT load and PUE, running units, and N, N+1 or 2N redundancy checked against Uptime Tier II to IV.
  • Emissions, fuel and enclosure — Region-specific emission packages for the US, Germany and other EU markets, fuel autonomy from tanks and bulk storage, the German bulk-storage permit threshold, enclosure fit by frame size and enclosure sound level.
  • Terms and the stated total — Paralleling switchgear, Premium service for continuous operation, lead time against the site date, a low-load review, discount authority by quantity, and a check that the quote total the agent states matches the total the validator computes.

It does not cover other Rolls-Royce product lines such as marine or aerospace propulsion, genset models outside the five modeled, real Rolls-Royce prices or terms, or applications other than data centers.

The AI sales agentBuilds the quote from the conversation and says it is ready.
Predictika validatorChecks the quote against the product and contract rules and returns every finding it has.

7 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 sales agent says

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

Proposing agent: AI sales 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 or Revenue Cloud) does with the same cases. It is a judgment about effort and reliability, not a claim that nothing can be built.

What the validator catchesTypical rules-based CPQPredictika
Genset count from load, rating and redundancyOnly with workarounds
The count comes from IT load, PUE, the kVA of the chosen rating class and the redundancy scheme, and the same count drives fuel, low-load review and paralleling; chaining those takes several rules that must run in order.
Each figure is computed once and every check reads them.
Emission eligibility by region and ratingOnly with workarounds
Eligibility is a matrix of region, rating class and package, and it grows with every new region or rule.
Region, rating and package are checked together on every call.
Fuel autonomyOnly with care
Runtime depends on the running units, burn rate, hours and the tanks plus bulk storage, so the check needs the computed unit count.
The need and the storage are computed and compared in the model.
Enclosure fit, noise and lead timeOnly with care
Frame size, container limits, sound level and lead time with aftertreatment are simple lookups that must be kept complete as options are added.
Enclosure, genset and options are read together and compared with the site limits.
Discount authority that depends on quantityOnly with care
A salesperson limit, a quantity-based maximum and an approval step need separate rules.
All thresholds are evaluated together and returned with the reason.
The total the agent statesOnly with workarounds
Comparing a stated total with a computed one means a separate rule per quote layout, and price rules rarely run on stated figures.
The stated quote total is a field in the request and is checked against the computed total.
Every problem returned in one answerOnly with workarounds
Rules tend to fire as the quote is saved, 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

Click to expand

Every rule is checked on every call, in whatever order the choices arrive, and one problem never hides another. Blocking findings stop the quote; approval findings need a person to sign off; everything else is computed.

RuleEffectWhat it checks
1computedRated kVA comes from the ISO 8528 rating class (ESP, DCC, PRP or COP); kW at 0.8 power factor
2computedFacility load is IT kW times the design PUE; running units are the load divided by unit kW, rounded up
3blockGensets quoted must cover the running units plus redundancy: N, N+1 or 2N
4blockUptime Tier III and IV cannot use the emergency standby (ESP) rating
5blockTier III needs at least N+1 and Tier IV needs 2N
6blockThe emission package must suit the region and the rating (US EPA, German TA Luft or Stage V, other EU)
7blockThe genset frame must fit the enclosure (20-cylinder units do not fit a 40 ft container)
8blockEnclosure sound level must not exceed the site limit
9blockTanks and bulk storage must cover the running units for the required hours
10needs approvalIn Germany, more than 100,000 L of bulk diesel needs a water-protection permit review
11blockMore than one genset requires paralleling switchgear
12blockContinuous (COP) operation requires the Premium service agreement
13blockLead time, longer with Stage V or Tier 4 aftertreatment, must fit the time available
14needs approvalFacility load below 30 percent of installed capacity needs an engineering review
15approval, then blockDiscount: 5 percent salesperson authority, approvable up to 7, 10 or 15 percent by quantity, beyond that blocked
16computedPrice roll-up from genset, aftertreatment, enclosure, fuel storage, paralleling and service
17blockThe quote total the agent states must match the validator total within EUR 1

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. Genset count: Six gensets

The agent's request. Six mtu DS3600 20V4000 gensets on the data-center continuous (DCC) rating with the Stage V package, in 45 ft sound-attenuated containers, with N+1 redundancy, 10,000 L sub-base tanks, 160,000 L of bulk storage and paralleling switchgear. Quote total EUR 10,555,000.00. Ready to send.

Verdict: Blocked.

  • Not enough gensets for the load and redundancy. Facility load 15600 kW needs 6 x 2880 kW units at DCC; 6 gensets are quoted but 7 are required.

2. Genset count: Seven gensets

The agent's request. Seven mtu DS3600 20V4000 gensets on the data-center continuous (DCC) rating with the Stage V package, in 45 ft sound-attenuated containers, with N+1 redundancy, 10,000 L sub-base tanks, 160,000 L of bulk storage and paralleling switchgear. Quote total EUR 12,245,000.00. Ready to send.

Verdict: Allowed.

3. Fuel autonomy: Sub-base tanks only

The agent's request. Three mtu DS2250 12V4000 gensets on the data-center continuous (DCC) rating, with no emission package required at this site, in 40 ft sound-attenuated containers, with N+1 redundancy, 10,000 L sub-base tanks and paralleling switchgear. Quote total EUR 3,140,000.00. Ready to send.

Verdict: Blocked.

  • Fuel storage too small for the required autonomy. 72 hours autonomy needs 67680 liters; tanks and bulk storage hold 30000 liters.

4. Fuel autonomy: Bulk storage added

The agent's request. Three mtu DS2250 12V4000 gensets on the data-center continuous (DCC) rating, with no emission package required at this site, in 40 ft sound-attenuated containers, with N+1 redundancy, 10,000 L sub-base tanks, 40,000 L of bulk storage and paralleling switchgear. Quote total EUR 3,220,000.00. Ready to send.

Verdict: Allowed.

5. The total the agent states: Stated total too low

The agent's request. One mtu DS2800 16V4000 genset on the data-center continuous (DCC) rating, no emission package required, in a 40 ft sound-attenuated container with a 10,000 L sub-base tank. Quote total EUR 1,083,000.00. Ready to send.

Verdict: Blocked.

  • Stated total differs from the validator total. The quote total the agent states (EUR 1083000.0) differs from the total the validator computes (EUR 1140000.0).

6. The total the agent states: Stated total corrected

The agent's request. One mtu DS2800 16V4000 genset on the data-center continuous (DCC) rating, no emission package required, in a 40 ft sound-attenuated container with a 10,000 L sub-base tank. Quote total EUR 1,140,000.00. Ready to send.

Verdict: Allowed.

7. Discount authority: Ten gensets, 12 percent discount

The agent's request. Ten mtu DS4000 20V4000 gensets on the data-center continuous (DCC) rating, in 45 ft sound-attenuated containers, with N+1 redundancy, 20,000 L sub-base tanks, paralleling switchgear and a 12 percent volume discount. Approval has been granted. Quote total EUR 14,529,000.00. Ready to send.

Verdict: Needs approval.

  • Discount needs approval. 12 percent discount is above the 5 percent a salesperson can give; approvable up to 15 percent.

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 sales agent prepares a quote for a diesel genset power plant for a fictional data center, in euros. Before it goes out, the agent sends it to Predictika’s validator, which checks it against the sizing, emissions, fuel, installation and pricing rules, compares the quote total the agent states with the one it computes, and returns a verdict with every reason.
  2. I will start with the first request by the agent.
  3. The agent says. Six mtu DS3600 20V4000 gensets on the data-center continuous (DCC) rating with the Stage V package, in 45 ft sound-attenuated containers. Quote total EUR 10,555,000.00. The rest of the request is on screen.
  4. Before it goes out, the agent calls the validator.
  5. Blocked, with one finding, all in one answer. Facility load 15600 kW needs 6 x 2880 kW units at DCC; 6 gensets are quoted but 7 are required.
  6. The validator returned one blocking finding 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. What this shows: for genset count from load, rating and redundancy, a rules-based CPQ can do this only with workarounds.
  8. Now the agent resubmits.
  9. The corrected request. It changes the quantity to 7 and the stated total as to 12,245,000 EUR.
  10. It calls the validator again.
  11. This time the validator clears it.
  12. The validator also returns the line-item quote. It starts with mtu DS3600 20V4000 genset and EU Stage V aftertreatment package; the other three lines are on screen.
  13. The earlier finding is cleared, and the validator's total, EUR 12,245,000.00, matches the quote.
  14. One problem caught in one call, fixed in one pass, and a clean quote. That is the value of Predictika's validator: catching mistakes before they reach the customer.