Best Place To BuildThe $45M Industrial AI Revolution | Daniel Raj David, CEO of Detect Technologies on BP2B S2 Ep.2
CHAPTERS
- 0:00 – 1:17
Detect Technologies’ mission: making industrial work safer with actionable AI
Daniel lays out Detect Technologies’ core mission: improving safety and efficiency in industrial environments using actionable AI. He frames IIT Madras’ ecosystem—professors, industry specialists, and especially fearless students—as the catalyst for building ambitious deep-tech solutions.
- •Mission focus: safer, better industrial workplaces through actionable AI
- •“Actionable” means insights that trigger real interventions, not just dashboards
- •IIT Madras ecosystem as an innovation flywheel
- •Students as the “X factor” who attempt what others assume is impossible
- 1:17 – 4:12
What Detect does in practice: an AI layer for cameras, sensors, and enterprise data
Daniel explains Detect’s product at a high level: a central AI layer that ingests visual feeds, time-series sensor signals, and even manual/ERP data. The outputs serve both frontline workers (real-time risk alerts) and executives (culture, safety, and efficiency metrics at scale).
- •Multimodal inputs: cameras, sensors, and “unstructured” operational data
- •Works for large plants and smaller fabrication yards
- •Delivers real-time frontline interventions plus management-level analytics
- •Enterprise-scale view of safety culture and operational efficiency
- 4:12 – 7:56
Why industrial safety needs automation: 24/7 risk detection and intervention
The conversation zooms into real factory scenarios and why traditional safety teams can’t monitor everything continuously. Daniel ties this to global fatality statistics and introduces OSHA as the rule framework that identifies leading indicators of serious incidents.
- •Industrial sites have known recurring risk patterns that lead to injuries/fatalities
- •Safety teams are limited in coverage; incidents often happen when they aren’t present
- •Need for always-on monitoring: 24/7/365 coverage
- •OSHA rules as the codified basis for identifying risky scenarios
- •Cost-effective scaling is essential for adoption
- 7:56 – 9:01
Beyond alerts: integrating AI with machines to prevent accidents (ports & suspended loads)
Daniel gives a concrete example from ports: detecting people under suspended container loads and automatically triggering machine stoppage and alarms. This chapter highlights Detect’s focus on closing the loop from detection to real-world intervention.
- •Example use case: ports, cranes, and suspended container loads
- •AI detects people in danger zones (e.g., under loads)
- •System can integrate back to machinery to stop operations
- •Goal is prevention—avoiding incidents and potential fatalities
- •Interventions designed for real-time on-site action
- 9:01 – 12:14
From preventive maintenance to high-temp sensing: expanding beyond video analytics
Detect’s scope includes sensor-driven preventive maintenance in complex industries like oil & gas. Daniel explains types of sensors (gas, pressure, ultrasonic) and a notable IITM-originated innovation (GUMPS) enabling high-temperature pipe monitoring with AI interpretation.
- •Multiple industrial sensor sources: gas, pressure, ultrasonic, thermal signals
- •IITM-originated GUMPS sensor: high-temperature signal generation via air-coupled approach
- •AI interprets sensor waveforms to predict defects and failures
- •Predictive insights: plan maintenance before breakdowns
- •Safety and efficiency are addressed together via early warning signals
- 12:14 – 15:11
The data moat: how Detect got training data for rare and hazardous events
A key challenge in industrial AI is building datasets for hazards you ideally want to avoid. Daniel explains how early deployments and partner sites enabled real-world data collection and labeling—forming a defensible advantage over purely synthetic approaches.
- •Training hazard models requires real-world data (synthetic wasn’t viable early on)
- •Origins in rule-based models (e.g., helmet detection) expanded to complex scenarios
- •Early industrial partners provided access and context to collect meaningful data
- •Complex hazards require multiple probabilistic models in parallel
- •Data access + labeling became a major competitive moat
- 15:11 – 17:14
OSHA risk categories Detect targets: PPE, line-of-fire, vehicles, and work-at-height
Daniel breaks down the kinds of OSHA-style safety rules Detect operationalizes. He illustrates multiple risk classes—from basic PPE compliance to high-severity “line of fire” incidents like walking under suspended loads and dangerous work-at-height behavior.
- •PPE compliance: helmets, masks, harness detection
- •Line-of-fire risks: people under suspended loads
- •Vehicle safety: overspeeding and improperly secured loads
- •Work at height: missing harnesses and unsafe scaffolding behavior
- •Non-compliance impacts: liability, fines, morale, and site culture
- 17:14 – 23:19
Origin story at IIT Madras: one email, one lab visit, and meeting the co-founder
Daniel recounts how his IIT journey began with curiosity about real-world impact, leading to an email to Prof. Krishnan Balasubramanian. That connection led him to a novel sensor project and to meeting Tarun Mishra—aligning industrial need with student-driven execution.
- •Mechanical dual degree (Intelligent Manufacturing), IITM 2012–2017
- •Desire to work on real-world tech, not “paper-only” outcomes
- •Prof. Krishnan Balasubramanian’s lab as the starting point
- •Sensor productization challenge created a path to industry exposure
- •Meeting Tarun Mishra connected industrial need with campus talent
- 23:19 – 26:18
Student-to-builder mode: industry trials, CFI talent, and balancing campus life
Daniel describes how his IIT years shifted heavily toward field trials and iterative learning with industrial partners, supported by CFI-style talent pools. He also reflects on maintaining friendships, sports, cultural activities, and the personal milestones that happened alongside intense building.
- •Heavy time spent in industry from second year onward for trials and deployments
- •CFI and campus groups supplied early engineering talent and momentum
- •Learning industrial realities shaped product direction and urgency
- •Still participated in table tennis (inter-IIT) and cultural activities
- •IIT as a flexible environment to explore multiple identities while building
- 26:18 – 28:42
Incorporation and focus: from many projects to a company (2016–2018)
The company incorporated in 2016 as graduation pressures and team scale made it “make or break.” Daniel describes an early phase of parallel experimentation—sensors, drones, analytics—before realizing the need to focus on where value truly sits and what can be outsourced.
- •Incorporated in 2016 to retain talent and commit beyond placements
- •Early scale: ~45–50 people involved on campus, largely unfunded
- •Worked across multiple tracks: sensors, drones as data capture, visual/thermal analytics
- •Mentors/investors helped shift thinking from invention-bias to product/value focus
- •Early industrial credibility: trusted as an “innovation lab” by major Indian enterprises
- 28:42 – 34:31
The pandemic pivot: hardware-agnostic AI SaaS, cloud vs edge, and scaling impact
COVID disrupted on-site deployments, forcing a pivotal strategic decision: keep the team, retrain roles, and move to hardware-agnostic AI SaaS running on existing infrastructure. Daniel also explains deployment architecture choices across cloud, on-prem, and edge—especially for low-connectivity sites.
- •Pre-2019: hardware + software; by 2019, massive labeled safety vision dataset growth
- •COVID blocked site access; revenue dropped and forced strategic clarity
- •Decision: no layoffs; retrain operational staff toward data/AI roles
- •Pivot: hardware-agnostic AI SaaS leveraging customer cameras and infrastructure
- •Architecture options: cloud/on-prem/edge; edge useful for remote/offshore, cloud for complex rule scale
- 34:31 – 46:26
Becoming global: first Shell contract, customer success standards, and regional operations
Detect’s globalization accelerated around 2019–2020, starting with Shell and expanding across industries and geographies. Daniel emphasizes that while the technology can scale, delivery and customer success must meet stringent local expectations—leading to regional entities and implementation teams worldwide.
- •First international contract: Shell; early deployments in the US around 2020
- •Expanded beyond oil & gas into construction, paper, mining, steel, and more
- •Global reputation flywheel: safety leaders share learnings across companies
- •Operational model: Chennai as product/R&D hub; regional teams/entities for delivery
- •India as a “toughest test bed” that stress-tests product quality and cost-effectiveness
- 46:26 – 48:30
Fundraising and validation: from seed to strategic investment (Shell Ventures)
Daniel outlines Detect’s funding journey across multiple rounds, highlighting how each stage aligned with scale and product maturity. A standout moment is Shell Ventures investing—validation from a top-tier customer-operator and reinforcement of Detect’s mission impact.
- •2017: ~$800K seed (angels + deep-tech funds like Accelor, Bharat Innovation Fund)
- •2018: ~$3.3M round led by Elevation
- •2021: ~$9M round with Accel/Elevation + initial Shell Ventures participation
- •2022: ~$25M round including Prosus, Accel, Elevation, Shell Ventures, and others
- •Strategic validation: Shell CEO award linked to “Goal Zero” (zero incidents)
- 48:30 – 1:00:08
Advice to builders: initiative, obsession, complementary co-founders, and IITM’s ecosystem advantage
Daniel answers how students should choose problems amid “problem overload,” stressing obsession and resilience over early certainty. He credits IIT Madras for labs, mentorship, incubation access, alumni networks, and CFI talent—ending with reflections on community and closing remarks.
- •Finding problems requires initiative and active exposure (especially for B2B)
- •“Problem of plenty” exists; winners are those obsessed enough to persist
- •Expect heavy skepticism; grit matters more than early approval
- •Co-founders must be complementary to sustain momentum through uncertainty
- •IITM advantage: labs (CNDE), incubation mentors, alumni access, CFI talent, and fearless students
