01
Client Profile & Background
A leading financial services company in Jakarta operates a critical data center housing over 200 servers supporting real-time transaction processing, core banking applications, and customer data storage. The facility relies on two 150-ton chillers with N+1 redundancy to maintain a server room temperature of 22±2°C and humidity at 45±5%. The IT team manages the environment manually through periodic walkthroughs and basic SNMP monitoring of the building management system (BMS), which lacks granularity and fails to provide early warnings for chiller degradation. With increasing regulatory scrutiny from Bank Indonesia and OJK, the company sought to enhance environmental monitoring to meet compliance standards and reduce operational risk.
The organization's infrastructure includes Dell PowerEdge servers, Cisco networking equipment, and VMware virtualization. The existing BMS only reported aggregate temperature and humidity at the room level, not per-rack or near critical equipment. The client needed a solution that could provide asset-level environmental visibility and integrate with their existing IT service management (ITSM) platform for automated incident response.
02
Technical Challenge
The primary technical challenge was the lack of real-time, granular environmental monitoring across the data center. The existing BMS had a polling interval of 5 minutes, which was insufficient to detect rapid temperature spikes caused by chiller fan failures or refrigerant leaks. Historical data showed that in the past year, two chiller malfunctions went undetected for over 30 minutes, leading to hot spots exceeding 35°C in specific racks. This resulted in three server crashes and one storage array failure, causing over 12 hours of downtime for critical applications. The financial impact was estimated at IDR 500 million per hour of downtime. Additionally, humidity fluctuations caused condensation on server components, leading to intermittent hardware errors that were difficult to diagnose.
The client required a monitoring system capable of detecting temperature variations as small as 0.5°C and humidity changes of 2% RH, with alerting latency under 30 seconds. The solution needed to cover 40 racks across two rows, including underfloor areas where water leaks from chiller condensation lines could go unnoticed. The existing network infrastructure had limited wired connectivity near the racks, so wireless sensors were preferred. The system also had to integrate with their existing hyperconverged infrastructure and Veeam backup environment for correlated event logging.
03
Implemented Solution
Intilogy deployed the MILESIGHT Chiller Server Room Monitoring solution using a combination of LoRaWAN sensors and a centralized gateway. Specifically, we installed 40 MILESIGHT EM500-SPT temperature and humidity sensors (one per rack), 10 MILESIGHT EM500-LPT water leak detection sensors placed under raised floors near chiller pipes, and 4 MILESIGHT EM500-CO2 sensors for air quality monitoring. These sensors communicate via LoRaWAN to a MILESIGHT UG65 gateway, which forwards data to a local server running MILESIGHT's IoT Hub software. The IoT Hub integrates with the client's existing server and storage environment, logging data to a Synology RS1221+ NAS for long-term retention and compliance.
The system was configured with threshold-based alerts: temperature >28°C, humidity <30% or >60%, and any water leak detection. Alerts are sent via email and SMS to the IT team within 10 seconds of threshold breach. Additionally, we integrated the IoT Hub with the client's Fortinet FortiGate firewall to enable secure remote access for monitoring. A custom dashboard was built using Grafana, pulling data from the IoT Hub's API, displaying real-time heat maps of rack temperatures and historical trends. The solution also includes automated escalation: if a sensor reading exceeds critical thresholds for more than 5 minutes, a ticket is automatically created in their ServiceNow ITSM platform.
04
Results & ROI
Post-implementation, the client achieved a 100% detection rate for environmental anomalies, with alert latency reduced from 5 minutes to under 10 seconds. Within the first three months, the system detected two water leaks from chiller condensation lines—one during a routine maintenance window and another during a power outage—allowing the team to respond before any equipment damage occurred. The average time to detect and respond to temperature excursions decreased from 30 minutes to 2 minutes, eliminating the risk of hot spots. As a result, server crashes due to thermal events dropped to zero, and the estimated cost avoidance from prevented downtime is IDR 6 billion annually.
The solution also provided compliance documentation for Bank Indonesia audits, with automated reports showing continuous environmental monitoring. The client's RPO for environmental data improved from 24 hours (manual logs) to real-time, with data retention of 1 year on the Synology NAS. The total cost of the solution, including sensors, gateway, and integration services, was under IDR 200 million, yielding a payback period of less than 4 months. Furthermore, the system's scalability allows the client to add sensors for new racks without additional infrastructure, supporting their planned 30% data center expansion over the next two years.
