Team FABLE 5, SIH26039. 93 references from our report and its addendum. 61 open the source directly, 18 open through the DOI, and 14 open a search for the exact title where the source has no stable address.
The single most important source for the gap analysis. Surveys fifteen coal-mine rescue robots over three decades and establishes that only ANDROS Wolverine V2 has a confirmed operational deployment record (seven.
The foundational field-report of mine rescue robot deployments and their failure modes - tether management, communications loss, and the robot itself becoming an obstacle. Corroborates Gap 1 independently and from the.
The reference design for an explosion-tolerant mine emergency vehicle: eight wheels, four cameras, CH4/CO/CO2/H2 sensing, nitrogen-pressurised sealed compartment, fibre-optic control. Used in Section 9.2 as the.
The landmark autonomous underground mapping result. Establishes that volumetric robotic mapping of an unmodified mine is achievable and defines the mass and cost scale at which it was achievable in 2003 - the baseline.
The Groundhog software architecture paper. Documents the 30 May 2003 run in which the robot autonomously descended 308 m into the abandoned Mathies mine and correctly decided to turn around at a blockage - the direct.
Describes the MSRBOTS system, which releases a 1 km fibre-optic cable from a passive autonomous spooling reel and needs two robots in series to reach 2 km. This is the clearest statement in the literature of the tether.
Field Robotics / arXiv:2207.04914, 2022. See also M. Tranzatto et al., "CERBERUS in the DARPA Subterranean Challenge," Science Robotics, vol. 7, no..
The co-top-scoring SubT entry, and the source of the design principle that a heterogeneous robot fleet should share one unified perception stack. Its tied 23-point score alongside CERBERUS is the evidence that neither.
The SLAM system behind the CSIRO Data61 result, and the system that proved most robust across sensing-degraded SubT environments. Its two-module structure - lidar-inertial odometry plus pose-graph optimisation - is the.
The consolidated cross-team survey of underground SLAM. Establishes that perceptual degradation is the dominant failure mode, that multi-modal sensing is not optional, and critically that loop closure - the mechanism.
The original frontier-exploration algorithm - move to the boundary between mapped free space and unknown space, and repeat. This is the exact algorithm DHWANI's EXPLORE mode implements, with 1 m frontier clustering and.
Demonstrates that inertial navigation can carry a position estimate through the short intervals in which UWB ranging is corrupted by multipath and NLOS in a coal mine - the exact fusion architecture DHWANI's EKF.
The governing Indian safety instrument. Supplies the methane thresholds used in the gas alarm logic (1.25 % vol for withdrawal of electricity in degree-three gassy seams), and Reg. 111(4)'s pillar-centre table, which.
The official Indian government safety chapter for the coal sector, giving the statutory framework, the DGMS reporting basis and the safety-initiative context in which this problem statement was issued. Used in Section.
The headline motivating statistic: 38 fatal accidents and 40 fatalities in Indian coal mines in 2024, against 38 fatal accidents in 2023. Establishes that the problem is current and not historical, and is the figure.
The official five-year series - 48 fatal accidents in 2020, 43 in 2021, 24 in 2022, 38 in 2023, 38 in 2024, and 226 deaths from fatal and serious accidents combined over 2020-2024. Provides the trend, not just a single.
The foundational theory: at UHF the rock and coal bounding a roadway act as low-loss dielectrics, so propagation is waveguide propagation rather than penetration, and the optimum frequency for minimum total loss is.
The NIOSH three-term model this project reimplemented and verified against Emslie's published Table V to two significant figures. Supplies the design attenuation figure of about 0.09 dB/m used in the link budget, the.
The Indian coalfield confirmation, from CIMFR Dhanbad, that a coal roadway supports hybrid HEM modes as a low-loss oversized dielectric waveguide and that total loss is minimal in the 450-1000 MHz range. This matters.
The key measured result at a frequency close to India's band: a path-loss exponent of n = 1.25 over 240 m in a 5 x 4 m drive (better than free space, n = 2.0), shadow fading sigma 4.5 dB, and 32 dB of knife-edge.
The independent measured corner loss: 30 dB per 90 degree turn (VVV) in a western high-seam coal mine, 40 dB at the Safety Research Coal Mine, and 48-55 dB with longitudinal antenna orientation. Corner loss is.
The authoritative source for receiver sensitivity in India's band (Table 10, RFS_L125_HF: -123 dBm at SF7 to -136 dBm at SF12, 125 kHz), for the airtime and bit-rate figures used in the link budget, and for the.
The only published hardware LoRa-mesh trial in a real mine found: 14 nodes at the Missouri S&T Experimental Mine achieving a 37.7 % packet delivery ratio with a topology-aware protocol and 46.8 % with plain ALOHA. This.
The best-performing thermal human-detection result located: a YOLOv4 model on NFPA 1801-qualified LWIR imagery achieving over 95 % precision at 30.1 fps in low-visibility smoky fire scenarios. Cited both as the standard.
Current state of the art in infrared-thermal fusion for rescue perception, and the source of the finding that no single infrared or thermal sensor simultaneously provides sufficient structural detail and reliable.
Establishes that mmWave FMCW radar can recover respiration and heart rate with relative errors around 1.3 % and 2.0 % against reference instruments, which is what makes DHWANI's second modality credible in principle..
Scientific Reports, vol. 8, art. 11551, 2018; and A. Ahmad, J. C. Roh, D. Wang, and A. Dubey, "Vital signs monitoring of multiple people using a FMCW.
Documents the limitations of the MOS methane sensors DHWANI carries as a low-cost demonstrator: high operating temperature, limited selectivity, long-term drift, slow response, and heavy cross-sensitivity to volatile.
Quantifies the cross-sensitivity problem specifically - traditional MOS sensors show significant cross-response to CO and ethanol when exposed to methane, leading to inaccurate measurement - and describes the.
The binding Indian instrument for electrical apparatus used in mines - which is the CEA regulation under the Electricity Act 2003, not DGMS's own regulation. Important for the compliance section because teams routinely.
Defines the actual certification route and cost driver for any Indian deployment: IS/IEC 60079-0 general, -1 flameproof Ex d, -11 intrinsic safety Ex ia, with coal mines as equipment Group I and BIS licensing mandatory..
Establishes that 865-867 MHz is India's delicensed short-range-device band at 500 mW e.r.p. with a 2.5 % duty cycle, and by omission that 433 MHz LoRa - widely used in international projects and in most hobby.
The architectural precedent for DHWANI's optional LLM planner: a language model proposes what is useful while a grounding layer constrains it to what is physically possible, with pre-trained low-level skills executing..
A NIOSH-funded dual-tracked mine rescue UGV with pan/tilt/zoom colour and thermal video, gas and temperature sensing, able to navigate 18 inches of water and 45-degree climbs. Three proof-of-concept units were built for.
The specification of the team's actual hardware platform: NVIDIA Jetson Nano, RPLIDAR A1, camera, IMU, wheel encoders, 3 x 18650 cells at 12.6 V, ROS Melodic with Gmapping/Hector/Karto/Cartographer support. Used in.
Melexis NV, MLX90640 32×24 IR Array Datasheet, Rev. 12,
Source for the stereo geometry in Section 8.5: 75 mm baseline, 73 degree horizontal field of view giving a 432 px focal length, global-shutter OV7251 sensors hardware-synchronised to plus-or-minus 100 microseconds. The.
Sensors, vol. 26, no. 12, art. 3701, 10 Jun. 2026, doi:
ations, 2023, No. CEA-PS-16/1/2021-CEI Division, notified 8 Jun.
Waveshare Electronics, JetRacer Professional Version ROS AI Kit , product documentation, 2024.
Frontiers in Robotics and AI, vol. 13, art. 1698570, 11 Feb. 2026, doi:
Applied Sciences, vol. 16, no. 18, art. 8986, 10 Sep. 2026, doi:
Tunnelling and Underground Space Technology, vol. 168, art. 107198, Feb. 2026, doi:
in Proc. 8th Int. Conf. on Mobile Systems, Applications, and Services (MobiSys '10), San Francisco, CA, USA, Jun. 2010, pp. 21 to 34, doi:.
IEEE Transactions on Mobile Computing, vol. 16, no. 2, pp. 341 to 354, Feb. 2017, doi:
IEEE Sensors Journal, vol. 22, no. 6, pp. 5417 to 5426, Mar. 2022, doi:
arXiv:2607.15807, 17 Jul. 2026 (rev. 3 Aug.
Applied Sciences, vol. 16, no. 12, art. 5861, Jun. 2026, doi: 10.3390/app16125861. (A distinct paper from [23] by the same group; [23] is the.
Strata Worldwide, Strata CommTrac , Wireless Communications and Tracking, product datasheet (MSHA IS approval nos. 23-A100008 and 23-A100012; PPL.
requency digital magnetic induction; link range up to 400 m, extendable with repeaters depending on rock type, ambient noise and antenna.
tion and Tracking Tutorial, CDC archive.
Journal of Field Robotics, vol. 43, no. 6, 19 Apr. 2026, doi:
arXiv:2509.26324, Sep. 2025 (rev. Mar.
Patterns, vol. 4, no. 9, art. 100804, Sep. 2023, doi:
in Proc. 3rd Conf. on Robot Learning (CoRL 2019), PMLR vol. 100, pp. 66 to 75, 2020;
Forensic Science International, vol. 165, no. 2 to 3, pp. 182 to 184, Jan. 2007, doi:
Slamtec, RPLIDAR A1 , 360° Laser Range Scanner, specifications (8,000 samples/s, 5.5 Hz nominal, 0.15 to 12 m, angular resolution ≤ 1°).
IEEE Transactions on Robotics, vol. 23, no. 1, pp. 34 to 46, Feb. 2007, doi:
in Advances in Neural Information Processing Systems 14 (NIPS 2001), Cambridge, MA, USA: MIT Press, 2002, pp. 713 to 720, doi:.
in Proc. IEEE/RSJ Int. Conf. Intelligent Robots and Systems (IROS), Vancouver, BC, Canada, 2017, pp. 5681 to 5686, doi:
in Proc. IEEE/RSJ Int. Conf. Intelligent Robots and Systems (IROS), Chicago, IL, USA, 2014, pp. 709 to 715, doi:
IEEE Transactions on Robotics and Automation, vol. 12, no. 6, pp. 869 to 880, Dec. 1996, doi:
IEEE Robotics & Automation Magazine, vol. 18, no. 4, pp. 80 to 92, Dec. 2011, doi:
in Intelligent Autonomous Systems 13 (IAS-13), Cham, Switzerland: Springer, 2015, pp. 335 to 348, doi:
re ARM Cortex-A57 at 1.43 GHz, 128-core Maxwell GPU, 4 GB LPDDR4, 472 GFLOPS, 5 W / 10 W modes).
8.2 at 1.7 GHz, 33 dense / 67 sparse INT8 TOPS, 7 to 25 W).
tte of India Extraordinary, Part II, Sec. 3(i), Sl. No. 142. Rules 3, 5, 11, 18, 19, 20, 21, 22, 24, 26, 27, 29 to 34; Schedules VI, VII.
clearance), Q.8, Q.11, Q.13, Q.15, Q.17, Q.18, Q.19, Q.20, Q.22, Q.25, Q.29.
Section II Gassiness, Statements 2.1 to 2.3; Section V Accidents, Statements 5.1, 5.3, 5.4.
Directorate General of Mines Safety, DGMS at a Glance (121 years), accident table 2008 to 2022, Crisis Management Plan, approval-policy.
in Annual Report 2018-19, subsidiary-wise rescue establishment table.
deaths in coal and lignite mines 2020 to 2024 (53, 51, 28, 41, 53), compiled by
16 Sep. 2025 (ex-gratia ₹15 lakh → ₹25 lakh; accident insurance ₹1 crore regular / ₹40 lakh contract).
16 Jun. 2020 , Explorer kit contents
Robot Price Index, Boston Dynamics Spot price page, verified Sep. 2026: USD 75,000 to 375,000; real configurations USD 150k to 375k from public.
ne]
99, Sep. 2026.
updated Jun.
IA international MSRP USD
PriceHistory.app, Luxonis OAK-D Lite Robotics Camera (India) , historical low ₹21,263.04, average ₹27,106, current listing ₹52,245. Retail.
0 breakout,
6.
27 Dec. 1975, Dhanbad , explosion followed by inundation; 375 killed; India's deadliest mining accident.
/ Mahabir Colliery capsule rescue, Raniganj, 13 to 16 Nov. 1989 , 71 trapped at ~350 ft; steel-capsule operation began 02:30 on 16 Nov., ran six.
(Ksan, East Jaintia Hills), 13 Dec. 2018 , 15 trapped at ~110 m, 5 escaped, search continued to 2 Mar. 2019.
17 Jan. 2019 , Indian Navy divers located a body at 50 to 65 m with the help of a remote-controlled probe.
(Umrangso, Dima Hasao), 6 Jan. 2025 , 9 trapped by inrush in a 300 ft rat-hole mine; Army, Navy, NDRF, SDRF, ONGC and Coal India deployed; all 9.
11 Jan. 2025.
recognised test houses CIMFR, ERTL, CPRI; Form B/C → scrutiny → test → QAR audit → field trial → Form K. Secondary industry source; the primary.
citing Coal India Limited Annual Report FY24 , 313 mines total: 131 underground, 168 opencast, 14 mixed; 34 discontinued underground mines.