AI in Operations
New AI tools optimize public transit networks and commercial construction site operations.
Anonymous
Key Points
1
The Salvador research drew on a dataset spanning about 700,000 passengers, roughly 2,000 vehicles, nearly 400 bus lines and almost 3,000 stops and stations.
2
Construction teams can choose among technologies such as RFID, BLE, UWB, GPS, equipment telematics and computer vision; the appropriate method depends on the needed accuracy and jobsite conditions, and the aim is not necessarily to track every object.
3
AI can be used to assess how safety equipment performs and wears over time, with the findings potentially informing safer product design.
AI tools are increasingly targeting transit and construction operations, promising to streamline decision-making and boost safety. In Salvador, Brazil, a research project deployed conversational agentic AI to analyze public transit data spanning 700,000 passengers across roughly 2,000 vehicles and 400 bus lines — helping managers spot demand patterns at specific stops and times PYMNTS. Meanwhile, construction is seeing broader AI adoption, though results remain mixed: a ServiceTitan survey found that 62% of commercial service contractors had piloted or deployed AI tools, yet only 15% of AI users reported significant positive impact with clear return on investment PYMNTS.
Salvador's AI-Powered Transit Network Optimization
Insight, a system developed in Salvador, uses agentic artificial intelligence to turn transportation data into actionable decisions for transit managers PYMNTS. The platform processes information from nearly 3,000 stops and stations, helping operators understand when and where demand peaks. This type of system reduces manual data review and lets managers respond faster to congestion and service gaps PYMNTS.
Construction AI Adoption Outpacing Proven Business Results
Construction teams are embracing AI quickly, but measurable success is lagging. Over 60% of contractors have tried AI tools, yet only 15% have seen clear financial returns PYMNTS. The gap suggests that deployment alone does not guarantee value — integration into daily workflows and proper problem selection are critical PYMNTS.
Common AI applications in construction include scheduling, dispatch, predictive maintenance, and safety oversight PYMNTS. Digital tools can analyze how protective equipment performs and wears, informing safer product design PYMNTS. Tracking tools and computer vision systems can monitor jobsites, though technology adoption remains uneven — many projects still rely on paper-based processes PYMNTS.
Edge AI Balances Speed With Reliability Challenges
Construction AI systems can connect jobsite equipment, materials, workers, and conditions to digital platforms — but connectivity is not guaranteed. Edge processing (running AI on-device rather than in the cloud) allows faster insights and works when cloud links drop Dev.to. On-device AI delivers faster responses, greater autonomy, and less cloud traffic Semi Engineering.
The trade-off is real: when cloud connections fail, queued data and commands must be validated before execution Dev.to. Not every queued decision is safe to run if conditions have changed Dev.to. Success depends on suitable sensors, networks, and context — technologies like RFID, BLE, UWB, GPS, and computer vision each suit different accuracy and jobsite needs PYMNTS.
Sentiment
Mostly PositiveNEGATIVENEUTRALPOSITIVE
Publishers
24
Articles
19
Reach
43
A
AnonymousAuthor
