Industrial data
Machine and shop-floor signals enter through approved sources, including the i4Gujarat / SIM ecosystem.


Industrial AI Centre of Excellence • Public progress update
A public progress update on the Industrial AI Centre of Excellence—from partnership announcement to a practical architecture, focused pilot portfolio, and implementation roadmap.
Joint solution frame
The proposed architecture connects industrial data, explainable AI, asset operations, and engineering governance—without removing human control from consequential decisions.
Machine and shop-floor signals enter through approved sources, including the i4Gujarat / SIM ecosystem.
Anomaly detection, rare-event reasoning, knowledge graphs, root-cause analysis, and explainable prediction.
Maximo turns insight into asset context, alerts, maintenance planning, and work-order workflows.
QRadar and IBM ELM strengthen security, requirements, testing, compliance, and lifecycle traceability.
Shared design principle: technology supports operators with timely evidence; accountable people remain in control.
Progress to date
The initiative has advanced from a public commitment to a defined technology model, a focused pilot portfolio, and an implementation pathway.
IBM, IAIRO, and the Government of Gujarat formalised and publicly outlined plans for an Industrial AI Centre of Excellence.
Read coverageJoint sessions mapped the roles of Maximo, SAMVIT, QRadar, and IBM Engineering Lifecycle Management.
Renewable asset intelligence and pharma quality control emerged as the strongest first-wave opportunities.
The team documented the time-series signals, operating ranges, anonymisation, and quality inputs required for modelling.
IBM and IAIRO consolidated a joint compute, storage, platform, and model-serving blueprint for consideration.
A stage-gated 90-day pathway now connects data readiness, technical integration, evidence review, and scale decisions.
Priority pilot portfolio
The first pilots are designed around measurable industrial value, explainable intelligence, and a clear path into operational workflows.
Connect equipment time-series data to SAMVIT reasoning and Maximo asset workflows to surface failures earlier and support faster maintenance decisions.
Use process and quality signals to detect abnormal behaviour, trace contributing conditions, and strengthen evidence for regulated decision-making.
Programme timeline
Each milestone below is grounded in documented activity. Future outcomes remain subject to partner, customer, and government approvals.
IAIRO shared its Industrial AI framing with IBM at the India Software Lab in GIFT City.
The partnership was formalised during the Vibrant Gujarat Regional Conference in Vadodara and entered the public domain through joint media engagement.
The team met Gujarat DST in person and received preliminary guidance on the infrastructure-support pathway.
The team met the MSME Commissioner and secured a route into the NPC and wider industry ecosystem for use-case and data discussions.
A support opportunity was identified for the Gujarat Industrial Policy 2026 chatbot alongside the industrial pilot track.
The teams mapped industrial asset data and SAMVIT intelligence into Maximo actions and operational workflows.
IAIRO prepared the infrastructure-request inputs while the teams assessed IBM ELM and QRadar capabilities for governance and security.
IAIRO requested sanitised industrial time-series data and defined the leading failure-mode contexts; production data delivery is not yet evidenced.
The teams aligned the platform roles, leading sectors, data options, and three-month proof objective.
The programme met Director ICT Kavita Shah (IAS) to advance the Industrial AI CoE roadmap.
The consolidated compute and storage plan was shared with Gujarat DST for guidance on next steps.
Brand Legal and ISA Legal discussions began for the required Maximo, QRadar, and ELM access; the supporting document is prepared and under review.
The modelling team defined the minimum IBM package: consistently sampled time-series readings and normal operating ranges for every feature.
A follow-up at the MSME office was planned to progress industrial data access, an NPC workshop, and use-case access; the outcome is not recorded in the reviewed material.
IAIRO shared the SIM understanding note and industrial challenge-problem set with IBM.
The weekly tracker continued to record the IBM Lab / Industrial AI CoE funding route as under discussion.
Current position
The programme has a defined solution frame and pilot pathway. The next gates depend on partner decisions, governed data, infrastructure guidance, and measured proof.
Wind assets and pharma quality remain the leading candidates. A final customer, accountable leads, and primary operational outcome are not yet evidenced.
The request and modelling schema are complete. Dataset delivery, consent, anonymisation, retention, and transfer controls still need confirmation.
The consolidated baseline was submitted to Gujarat DST on 2 September. Approval, allocation, and provisioning are not yet evidenced.
No live Maximo–SAMVIT integration, recorded demonstration, measured result, or acceptance decision is yet evidenced.
As of the 14 September tracker review, the funder, amount, commitment, and delivery route remained unconfirmed.
IBM legal review for Maximo, QRadar, and ELM access has started. Partners must also settle IP, data, security, hosting, support, and commercial responsibilities.
Evidence boundary: completed means the activity or document is recorded in the reviewed material. It does not imply funding, infrastructure, production deployment, or customer acceptance unless explicitly stated.
Next phase
The programme now moves from design to a tightly scoped proof: one customer context, one approved dataset, one end-to-end workflow, and measurable results.